|
Forever_Peace, I accidentally left plotly-latest.min.js in the root of my fork. I forgot to delete it when I issued a Pull Request. You can safely delete it.
|
#
?
Mar 31, 2016 01:18
|
|
|
|
| # ? Apr 19, 2024 22:17 |
|
|
Rabid Snake posted:Forever_Peace, I accidentally left plotly-latest.min.js in the root of my fork. I forgot to delete it when I issued a Pull Request. You can safely delete it. Done! And thanks for taking care of that - I had forgotten I hadn't finished fixing the graphics bug in the rest of the old apps! While we're at it, you have any thoughts about the other warning that tends to pop up with these? Update for the rest of you: I've finished the analysis for chapter five, but am having a hell of a time writing it. "Situational factors" turns out to be a big complicated topic. We may have to break it in two. While I figure it out, anybody have any easy Quick Hit requests so I can keep content coming to the thread?
|
|
#
?
Mar 31, 2016 13:26
|
|
|
Forever_Peace posted:Done! And thanks for taking care of that - I had forgotten I hadn't finished fixing the graphics bug in the rest of the old apps! While we're at it, you have any thoughts about the other warning that tends to pop up with these? I would really like to see a breakdown of Lamar Miller. He was one of the top Free Agent RB's this offseason and the general goon consensus is that Miami criminally under-utilized him. Do the numbers agree? What type of runner is Miller?
|
|
#
?
Mar 31, 2016 16:30
|
|
|
Forever_Peace posted:Done! And thanks for taking care of that - I had forgotten I hadn't finished fixing the graphics bug in the rest of the old apps! While we're at it, you have any thoughts about the other warning that tends to pop up with these? How do you feel about looking at QB runs and scrambles? I'm curious to see how, say, Wilson compares to Lynch, or how Brady compares to Kuhn (my hunch is that they're the premiere short-yardage guys at their positions.)
|
|
#
?
Apr 1, 2016 00:12
|
|
|
Adrenalist posted:How do you feel about looking at QB runs and scrambles? I'm curious to see how, say, Wilson compares to Lynch, or how Brady compares to Kuhn (my hunch is that they're the premiere short-yardage guys at their positions.) Brady is hilariously good at getting a foot.
|
|
#
?
Apr 1, 2016 16:15
|
|
|
Adrenalist posted:How do you feel about looking at QB runs and scrambles? I'm curious to see how, say, Wilson compares to Lynch, or how Brady compares to Kuhn (my hunch is that they're the premiere short-yardage guys at their positions.) And try to throw Tebow scrambles into the comparison.
|
|
#
?
Apr 2, 2016 02:58
|
|
|
MrSargent posted:I would really like to see a breakdown of Lamar Miller. He was one of the top Free Agent RB's this offseason and the general goon consensus is that Miami criminally under-utilized him. Do the numbers agree? What type of runner is Miller? Quick Hit: It's Miller Time Lamar Miller was one of the goon favorites among the free-agent running backs this year. Among the only University of Miami players drafted by the Dolphins in nearly a decade (along with Olivier Vernon in the preceding round), Miller has put in three solid, healthy years as the starting running back, missing no games and averaging 1174 scrimmage yards per season. Thanks to some serious speed (Miller had the best RB 40 time of 2012 and would have been top 3 in every year since), Miller has quietly been an efficiency king at Miami, posting a career YPC of 4.59 and a respectable success rate. But bizarrely, Miller's 2015 box stats seemed somewhat pedestrian, thanks to a volume of only about 12 carries per game. A number of people started speculating that Ajayi was seen as the future of the backfield, or perhaps that the franchise didn't trust the run game at all. So what do we know know about Lamar Miller as he makes his jump to the Texans? Well, to start with, he has clearly ran like an above-average back.  Though his career 4.59 YPC might not seem too far from the 4.18 league-average, 638 carries is a pretty sizable sample, meaning that we can be pretty confident that Miller has been an above-average runner during his time the league. In fact, a league-average running back would only produce a YPC that high given 638 carries only about 5% of the time.  That doesn't just meet the "yeah he's probably good" standard of evidence. It just about meets the standard of evidence that scientists use when publishing professional research. One thing we can look at with our tools is how quickly Miller is able to distinguish himself. We'd only need to see about a 100 carries of each to be able to conclude that Miller is almost certainly a better runner than Trent Richardson:  But at the other extreme, Miller actually keeps pace pretty well when stacked up against Jamaal Charles, the best runner in our database. We'd need over five hundred carries from each before we'd be able to comfortably conclude the Charles is clearly the better runner.  In terms of actual comparables, our player matcher comes through yet again with a #1 match that I really like:  JStew put in a lot of years with similarly low workloads (though by way of a shared backfield), producing with similar efficiency and with a similar penchant for bursts of solid gains. If Miller continues in the NFL as a healthier Stewart, he could have a great career. And importantly, his successes are probably not attributable to some monster offensive line. Some of the lines he's worked behind over the past few years have actually been mediocre at best (if not outright deficient). But in addition to the fact that his primary strength is in breaking mid-to-long runs at an above-average clip (where the offensive line contributes most to the RB's success through the contested yards close to the line of scrimmage), Miller has also pretty decisively outrun his teammates:  Notably, there are no discernible weaknesses to his rushing output. If he was producing like this with 15 carries a game instead of 12, we'd consider him one of the game's top grinders, and if he was able to maintain that output as a feature back with 20+ carries per game, he'd be a regular contender for the rushing title. I think beat reporter James Walker summarized Miller well: quote:The mystery with Miller has always been can he carry the ball 20-25 times a game? Miami's coaching staff determined early in his career -- maybe his first or second year -- that Miller was best at 10-15 carries and often stuck to that throughout his tenure. Miller even gained about 10-12 pounds of muscle last year and things mostly stayed the same. That sounds pretty spot on. One complaint I still hear about Miller is that he's a "good 2-down back", implying perhaps that he is a liability in the passing game or a poor pass-catcher. Now, this may have been true when Miller came into the league, but both pass protection and yards through the air were the two primary areas where Miller has improved in leaps and bounds during his time with the Dolphins. Miller is now at least a decent pass blocker, and his share of the passing game has increased steadily every year. In 2015, he had about as many catches as Gio Bernard and nearly as many as Charlie Sims (two guys known for their role as pass-catchers). As a starter, his air yards increased by 60% between 2013 and 2014, then by an additional 45% between 2014 and 2015. Last season, an extra 1/4 catch for 5 yards per game would have given him the same per-game output as Matt Forte. There's also little evidence that his production has been limited on 3rd downs. He carries the rock on 3rd downs at pretty much the same rate we would expect given the league-wide running game as a whole:  However, as Walker noted, the major concern for Miller is whether he ever can shoulder the type of volume that would put him in the same conversation as Demarco Murray and Frank Gore. First, the Dolphins just aren't a run-heavy team. Last year, they had by far the fewest rushing attempts in the NFL. But even as a proportion of his team's run game, Miller has never been one of the volume Grinders. He tends to carry about 60-65% of the RB touches as a starter, with a career usage that's actually pretty modest:  It will be interesting to see if the Texans keep him on the field more. But even if they give him exactly the same share of the workload as he's used to, Miller will be getting significantly more carries: The Texans have been top-5 in rushing attempts for two consecutive years now (I mean, with their quarterback situation... wouldn't you?). If Miller had been carrying exactly the same portion of the run game last year on the Texans instead of the Dolphins, he'd have 70 additional carries and would have had the 5th highest RB volume in football. Start dialing up his share of the carries alongside the increase in team carries, and you're looking at the potential for some serious numbers. But it's quite possible that Miller just can't handle that kind of volume. Honestly, we just don't know, because the Dolphins have never used him in that manner. And 2015 didn't do him any favors. On the ground, it was his worst year to date. And not just in terms of the volume stats - it shows up in the rate stats and run distributions as well. Miller was pretty consistent in his first three years, posting a good 50-carry rookie year, a decent sophomore year as the team's starter, and a great 2014 as starter.   But then he backslid.   That looks suspiciously like slowing down. This is apparent in the comparison to his teammates. In 2014, he was head and shoulders the best runner on the team:  And in 2015, that clear dominance started to fade:  Sure, the composition of the backfield changed between years, but I'm not sure we can just brush this off by saying that "maybe Ajayi and Jonas Gray are really good???". However, even 2015 had it's bright spots. Though his ability to break through the contested yards clearly suffered, he still managed to tear out a few big chunks of yards here and there.  This does look somewhat like the product of a good runner behind a bad line. Given better blocking, I would be cautiously optimistic of more years like 2014. In sum, I think there's a lot to like with Miller. He's tough, efficient, fast, versatile, and has improved significantly as an air threat and pass blocker. He's shown some considerable talent on the field as a runner, and the Texans should provide plenty of opportunity for him to carry the rock. But there is plenty of risk for Houston as well. Miller has never handled a consistent volume, so it's unclear whether his record of efficiency and durability will continue. Further, 2015 really was a down year for him, and not just because he was underutilized. The carries he did have just didn't look like his usual self. It's possible he's already lost a step. But ultimately, I probably come down on the optimistic side. Though the offense as a whole leaves a bit to be desired, new acquisition Jeff Allen is a monster run-blocking guard, and the team's previous best run blocker Derek Newton just signed a 5-year extension at tackle. If the unit clicks like it should, I think Miller will see enough room to break out plenty of his characteristic bursts. I like the move for Houston. As sad as it is to see Foster walk, I think Miller is a really smart choice for a successor.
|
|
#
?
Apr 2, 2016 22:20
|
|
|
Thank you so much FP! That was a fantastic read and confirmed a lot of my thoughts on Miller. I really can't wait to see what he does on the Texans next year.
|
|
#
?
Apr 6, 2016 21:06
|
|
|
I'm pretty happy about having Miller on my dynasty team now. After the draft last year it seemed like I had badly overpaid, and it was agony watching Miami chronically underutilize him (and chronically lose games doing so). The Texans aren't exactly a premiere football team but given their QB situation I like the prospects for a better Miller year, and for him to work up to a bell cow utilization. Plus he's still quite young.
|
|
#
?
Apr 9, 2016 07:40
|
|
|
Leperflesh posted:I'm pretty happy about having Miller on my dynasty team now. After the draft last year it seemed like I had badly overpaid, and it was agony watching Miami chronically underutilize him (and chronically lose games doing so). The Texans aren't exactly a premiere football team but given their QB situation I like the prospects for a better Miller year, and for him to work up to a bell cow utilization. Plus he's still quite young. Yeah I lamented drafting Miller for my dynasty league most of last season. Now I might actually have an RB1! I love the JStew comparison, big fan of JStew.
|
|
#
?
Apr 11, 2016 15:24
|
|
|
MacheteZombie posted:Yeah I lamented drafting Miller for my dynasty league most of last season. Now I might actually have an RB1! I love the JStew comparison, big fan of JStew. You had one last year  Miller owns, even though I traded him away in my main league, I'm hoping he blows up like a durable Foster would've the past few years.
|
|
#
?
Apr 12, 2016 21:03
|
|
|
Holy crap I guess I didn't realize he finished 4th in RBs in .5 ppr last season. Wow.
|
|
#
?
Apr 12, 2016 21:19
|
|
|
MacheteZombie posted:Holy crap I guess I didn't realize he finished 4th in RBs in .5 ppr last season. Wow. Yeah. The problem was that he was used pretty inconsistently and you never knew any given week whether he was going to get you decent points or not. Also Miami generally sucked anyway. But I started him pretty much straight through the season and mostly didn't regret it. He's young, talented, and hopefully now on a team where he'll not just be the bellcow, but be the bellcow on a team that understands that running the ball is an OK way to play football sometimes.
|
|
#
?
Apr 22, 2016 23:09
|
|
|
Any quick hits on draft prospects, or is ncaa data not as easily available?
|
|
#
?
Apr 29, 2016 01:02
|
|
|
effectual posted:Any quick hits on draft prospects, or is ncaa data not as easily available? I pm'd Forever_Peace a while ago about this, and even went hunting for NCAA data, but unfortunately there's no database similar to NFLDB that I found. I did find a bunch of compiled spreadsheets but I believe it was missing the last few sheets and I don't think the data was in a format usable for this. I hope FP found something that would work, but from the googling I did it seems difficult to obtain play by play data for college games for various reasons.
|
|
#
?
Apr 29, 2016 18:50
|
|
|
MacheteZombie posted:I pm'd Forever_Peace a while ago about this, and even went hunting for NCAA data, but unfortunately there's no database similar to NFLDB that I found. I did find a bunch of compiled spreadsheets but I believe it was missing the last few sheets and I don't think the data was in a format usable for this. MacheteZombie is right - there's no clean database that I could find for college ball. No quick hits for the draft - sorry! =( But FootballOutsiders just published their "BackCAST" predictors which look fun.
|
|
#
?
Apr 30, 2016 15:22
|
|
|
Next step to developing a predictive model should be trying to separate RB performance from the surrounding cast, but I don't know if you have the database for that on hand. Going to give it a shot? Or perhaps a confidence based model that can quantify similarity year-to-year in surrounding cast, and then forecast RB performance with some variable margins for error, or distributions of results? Basically, you have dozens of readers wanting to know who to draft in fantasy this season (that's okay to talk about in broad generalities, here, I hope?). But you already knew that
|
|
#
?
May 4, 2016 03:17
|
|
|
The best way would be to identify RBs who got traded mid-season. Unfortunately I suspect there aren't enough to build a statistically useful and valid data set. e. and that also ignores that a freshly-traded RB is coming into a new playbook and will probably not be used the same way as a team veteran runner. So instead I think we're stuck with trying to quantify aspects of the surrounding team - the OL's ability to open holes, the team's emphasis on passing vs. running (which affects defenses tendency to position for run stuffing vs. pass coverage), the qualities of the QB (has the same effect), types and frequencies of different run plays (QB option plays vs. designed runs, for example), game situations (behind vs. ahead in the 4th quarter for example), and maybe even field conditions (outdoor winter live grass vs. indoor may affect run success?) etc. A daunting job. There are so many variables it'll probably never be something where you can boil everything down to a "how hard is it for any random runner to average 4.5 yards with this team" number that you could be very confident in. Leperflesh fucked around with this message at 19:00 on May 4, 2016 |
|
#
?
May 4, 2016 18:57
|
|
|
Yeah if I had to I would just resort to PFF grades I think.
|
|
#
?
May 4, 2016 20:05
|
|
|
I read through this all in one go. Really enjoyed it, and learned a lot. I currently don't have a computer, but playing with your Apps is now as much of an incentive to get a new one as the various games I haven't been able to play.
|
|
#
?
May 15, 2016 08:51
|
|
|
Sorry if this has been asked before didn't see it, is there a way to possibly request a short summary of a player that played close to but a little over 6 years ago? Was gonna ask about Shaun Alexander
|
|
#
?
May 15, 2016 10:07
|
|
|
The Outliers: The Greatest Single-Game Rushing Performances of 2015 Greatness can be hard to measure. So much of what makes a player great is wrapped up in the details and the circumstances: unforgettable moments, epic comebacks, and superhuman displays of skill. But I got to thinking recently, how would one go about trying to measure a truly great game? Plenty of folks have tried their hand at this before. They might add up yards or touchdowns, highlight the toughness to play through injury or the leadership to carry a team, or note astonishing efficiency. But I wanted to add something different to the conversation. One thing we discussed a lot in chapter 4 was the role of random chance at small sample sizes, such as those typically seen for single games. Mike Anderson has one of the top-10 single game rushing totals of all time, and Jerome Harrison is #3, but neither is considered one of the all-time greats (though Harrison is a pretty crazy story for other reasons). So I decided to phrase the question in a particular way: what is the most improbable game. What’s the game that a league-average player is least likely to have?  Technical math below. Scroll down if you just want the list. Here’s how I went about it. 1) Estimate the league-average probability of each run. At the core, we expect that the probability of getting (for example) a 5-yard run on any given play will be pretty close to the proportion of the time that players have gotten 5-yard runs on any given play in the past. However, once you get up to 53-yard runs, 54 yard runs, 55 yard runs etc., the data gets pretty sparse and the estimates become unstable. So we smooth over our estimates a bit, just like we did in chapter 1 to get this:  In practice, this kind of smoothing just means that our guess for how probable a 73-yard run should be is also influenced by the rate that people get 71, 72, 74, and 75-yard runs. (Specifically, we used guassian kernel density estimation with a bandwidth around 3.5) Once we had stable estimates for the longer distances, we found the probability of every single run distance (specifically, by sampling the density at each point and dividing these densities by the sum of all densities to make them add to 1). 2) We only care about improbable runs that are “good”. A game where a player goes for -5, -4, -4, and 10 might be more improbable than a game where a player goes for 2, 3, 3, 10, but it’s certainly not “greater”. So we make a “bin” for all of our bad improbable runs. Specifically, we find the most probable run (i.e. the mode of the run distribution), which as you might recall from chapter 1, is 2 yards. Then, we make a “bin” where we can lump together all of our “bad” improbable runs – the runs that go for less than 2 yards. It turns out that about 32.19% of all runs go for 2 yards or less. We care about all of the “good” improbable runs that make it past this “loser group” of runs. 3) We find the odds that a league-average running back could have produced each game over the past six seasons. We can do this simple using the multiplication rule. If the probability of getting “tails” on a coin flip is 0.5 (i.e. 50%), the probability of getting two “tails” in a row is 0.5*0.5 = 0.25 (i.e. 25%). We basically do the same thing, but for the probabilities of each individual run in the game. To go from probability to “odds”, we basically take the ratio between the probability that a league-average back does have exactly that game vs. the probability that they don’t have that game. E.g. a 0.4 probability (40%) is 2/3 odds (sometimes also written as 1:1.5 odds). 4) We find the odds that a league-average running back would have just produced “loser group” runs all game (i.e. the big group of runs we made of 2 yards or less). These “meh” runs should occur about 32.19% of the time, so if a player took 20 carries, we can calculate the probability that all of them were “meh” runs. A game of “meh” runs is more common by far than any other specific outcome, because we combined a bunch of runs all into this one loser group. 5) We take the ratio of these two odds. The odds of the game they had / the odds of a “meh” game given that many carries. Also known as the “odds ratio”. This is a way for us to estimate the special sauce that a particular player brought to the game above and beyond a game of league-average “meh” runs. 6) Add a fudge term. Because we want to compare 10-carry games to 35-carry games, I added a stiffer penalty for the number of carries. I just multiplied our odds ratio by number of carries. Technical poo poo over #5 - David Johnson @ PHI, week 15 The red-hot 11-2 Cardinals looked to clinch the NFC West for the first time since 2009. Rookie David Johnson had exploded onto the scene two weeks prior after Chris Johnson fractured his tibia, rolling over the Rams for 120 yards from scrimmage and following it up with an additional 120 yards the following week against the Vikings. David Johnson was showing power and balance in spades, with the arrow clearly pointing up, but I don’t think anybody could have anticipated what happened next. Despite missing much of the second quarter with a knee injury, DJ gashed an atrocious Eagles run defense for 187 yards and 3 touchdowns, adding an additional 42 yards through the air. But what was truly impressive here was how those yards were gained. Of his 29 attempts, 21 went for more than two yards, and two of those “short runs” were touchdowns from the 1 yard line where he could not have gone further anyways. He took the expected run distribution and blew it up, gaining considerable ground on nearly every play:  (DJ in red, league average in blue) When the dust settled, he had become the first player in team history to rush for 180 or more yards and score three touchdowns in a game, set the franchise rookie record for touchdowns in a season, clinched the divisional title, and pushed his team to a franchise record for regular season wins. #4 – Carlos Hyde vs. MIN, week 1 https://twitter.com/SBNation/status/643629524681400320/ Hyde had a lot of question marks coming into his sophomore year. Long-time stalwart Gore had left for the Colts, and it wasn’t clear whether Hyde had it in him to fill Gore’s shoes after a tepid 333-yard rookie year. Hyde responded to critics with a virtuoso performance in consistency.  (Hyde in red, league average in blue)  On average, only about a quarter of all runs make it past five yards. Half of Hyde’s 26 carries went seven yards or more. Six runs went for 10 yards or more. With a longest run of only 18 yards, I think a lot of folks are quick to overlook just how impressive this game was because of the relatively modest final tally of 168 yards and two touchdowns. But our analysis suggests that this game was nearly an order of magnitude better than David Johnson’s at #5. #3 – Chris Ivory @ MIA, week 4 The Jets go to London and unleash a monster onto Wembley stadium. I couldn’t count the number of time Ivory ran right the gently caress at Suh and ended up winning. He ran with grit and power, and it showed with a remarkable number of attempts that completely gashed a shellshocked Miami defense.  (Ivory in red, league average in blue) He pulled off 9+ yard runs at double the average rate on his way to a 29-carry 166-yard performance. I personally liked Hyde’s relentless week 1 a little better, but it’s notable that Ivory had an uncanny knack of picking up the one extra yard that this analysis really seems to like – the runs that Hyde took 5, Ivory took 6, and the runs that Hyde took 7-8, Ivory took 9. #2 – Doug Martin @ PHI, week 11 The highlight-reel game of the list. After two down years, Martin finally returned to form, posting a mid-season string of three consecutive 100-yard games. But this was the jewel of his season. He finished the game with 235 yards in 27 attempts. What the 8.7 YPC doesn’t show, however, is that this game was one of the boom-bustiest of year.  (Martin in red, league average in blue) Aside from the impressive total, this game is notable for a) a ludicrous number of 3-yard runs, and b) a trifecta of barnstormers, going for 27 yards, 58 yards, and 84 yards, none of which scored a touchdown. It may not look like much compared to Hyde’s astonishing consistency, but those big runs are highly improbable to all happen in the same game. In a typical 200-carry season, we’d only expect five runs total that go for 20+ yards. Martin had three in a single game. Martin broke out two 40+ yard runs in a single half, while only two other running backs had more than two 40+ yard runs in all of 2015: Adrian Peterson and Todd Gurley. And for that reason alone, this touchdown-less game of home runs just barely beats out Ivory for the #2 spot. And the NFL noticed, giving him “ground player of the week” for his efforts. But the NFL hosed up. Because that same week, somebody else was busy setting the most astonishing rushing game of 2015 by far. #1 – Thomas Rawls vs SF, week 11  This little 5’9 undrafted rookie was asked to fill in for Beast Mode on a number of occasions this year. Week 3, he dropped a cool hundred yards on Chicago in his first start in the NFL, then took the Bengels to the shed for 169 yards two weeks later before Lynch came back. Fast forward six weeks and Lynch is out again with an unexpected abdominal injury, leaving Rawls to again pinch-hit for the big man. Rawls made the most of the opportunity. I want to underline how batshit this game was. Seattle is not known for their offensive line, to put it mildly. Marshawn Lynch has never broken 160 rushing yards, and has never broken 200 yards from scrimmage. Rawls decimated the Niners for 209 rushing yards (and a touchdown) on 30 carries, with an additional 46 yards and a touchdown through the air. It’s the most rushing yards ever by a rookie in franchise history. It’s the first time a Seahawk has run for 200 yards since Shaun Alexander. It’s also the first time in NFL history that any rookie has ever gained 250 yards from scrimmage with both a rushing and receiving touchdown. Now get this: the algorithm I’m using doesn’t know any of this. It doesn’t know about the shaky line or the receiving yards or the touchdowns or the records. This is why my algorithm likes this game:  (Rawls in red, league average in blue) Rawls was just untouchable that day. Consider this: Adrian Peterson had by far the most 10+ yard runs this season with an average of 2.68 big runs per game (Martin was 2nd with 2.06 big runs per game). Rawls had seven in this one game alone. Here’s the (sorted) game log: -4, 0, 0, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 5, 5, 5, 5, 7, 8, 8, 9, 10, 11, 17, 18, 20, 22, 30 Rawls was just spectacular. The metric thinks that this game was the greatest of the season by four orders of magnitude, towering above everything else. Rawls could be pretty good, y’all. Honorable Mentions: Best Game under 25 carries: Matt Forte, week 1. Best Game under 20 carries: Todd Gurley, week 4. Best Game under 15 carries: Lamar Miller, week 7. Best Game under 10 carries: Karlos Williams, week 9. (game log: 1, 3, 5, 5, 10, 11, 12, 25, 38) Best Game by a Player Over 30: Rashad Jennings, week 17. Most Appearances in the top 20: Adrian Peterson with four, weeks 2, 9, 10, and 12. Worst Game of the Year: Joique Bell, week 3. (game log: -1, 0, 0, 0, 0, 1, 1, 1, 1, 3) Forever_Peace fucked around with this message at 18:52 on May 15, 2016 |
|
#
?
May 15, 2016 18:16
|
|
|
This is a good post, and not just because of who's at #1. However: Forever_Peace posted:the probability that they [/i]don’t[/i] have that game.
|
|
#
?
May 15, 2016 18:44
|
|
|
whypick1 posted:This is a good post, and not just because of who's at #1. Thanks for catching. Fixed! Now I need to decide if this counts as a "Quick Hit" or should be listed as something else. "Interlude"? Chapter 5a (where we go through 5f)?
|
|
#
?
May 15, 2016 19:01
|
|
|
Without knocking Rawls at all: I don't think you can just ignore game situation. At that point in the season, the Seahawks were at 4 and 5, tied with St. Louis and behind a dominating 7-2 Arizona. The 49ers, on the other hand, were at 3-6. Despite being only two wins out of second place in their division, the 49ers had all but given up on the entire season. Worse, the niners had just lost against the Seahawks at home three weeks prior, and were now playing at Seattle, where the Hawks have a distinctive advantage. True, they'd eked out a slim win against the Falcons, who were otherwise having a pretty good season, but all of their division games had been blowouts and the team was struggling mightily to keep their offense on the field at all. The niners' defense had been absolutely gutted in the 2015 offseason. Kaepernick was benched; this game was Blaine Gabbert's second start of the season. And Russel Wilson completed 24 passes of 29 attempts, preventing the niners from focusing on a run defense. Basically, the Hawks were in a position to compete for the wildcard spot, knew they could beat the niners, and firing on all cylinders. The niners were demoralized, shattered defensively, badly coached, and despite not being out of the running for the playoffs, appeared to have basically given up. Blaine Gabbert actually handed in a pretty decent performance, but the rest of the team did not. It was a perfect situation for a Seattle RB to tear rear end, and Rawls showed up. He deserves full credit for his performance, but situation made it possible. I think you can look at the other top games and see some differences along those lines, too. Miami's defense in london was probably easier to run against than the league norm, for example.
|
|
#
?
May 15, 2016 19:34
|
|
|
Forever_Peace posted:
Dat gap is sexier than a swimsuit model's thigh gap, and it ain't even close.
|
|
#
?
May 15, 2016 19:42
|
|
|
Scionix posted:Sorry if this has been asked before didn't see it, is there a way to possibly request a short summary of a player that played close to but a little over 6 years ago? Was gonna ask about Shaun Alexander I'm also super curious about him, for those few years he was everywhere, but people maintained that he was bad. I'd love to see the tale of the stats.
|
|
#
?
May 15, 2016 20:07
|
|
|
McSpankWich posted:I'm also super curious about him, for those few years he was everywhere, but people maintained that he was bad. I'd love to see the tale of the stats. He ran behind one of the more dominant lines in NFL history. Which makes me wonder how Priest Holmes compares in a similar situation if Forever_Peace wants to check it out.
|
|
#
?
May 15, 2016 21:06
|
|
|
Leperflesh posted:Without knocking Rawls at all: I don't think you can just ignore game situation. Maybe! I completely agree that there was something specific about this Seahawks team meeting this SF team at this particular time that helped contribute to the monster game from Rawls. But that's kind of the point - this is a list of games that were special. Games that you just don't expect to see. I'm OK if some of that is attributable to a bad defense (PHI gave up two of the top 10 greatest rushing games this year) or a motivated team or whatever. Though the thing about emphasizing these narratives is that it can be hard to pin down the most important ones sometimes. It's true that Rawls was working with the "Tom Cable special" for the offensive line, but it's also true that SF were looking dire and were possibly demoralized. Does it change things to know that Lynch had the second best game against SF, from four weeks earlier? Perhaps the narrative that matters here is that Seattle just has SF's number right now? Maybe that was supposed to be Lynch going for 250 scrimmage yards? Or how about knowing that the week after Rawls' went haywire, SF stifled the Arizona Johnsons for 38 combined yards on 20 carries? The narrative here would be that SEA went bananas over a defense that Arizona could barely dent on the ground just one week later, indicating how much stronger they were as an offense. Or maybe the same thing actually means that SF was spurred into trying again after getting humiliated? Does it matter that most of Rawl's best runs happened in the 4th quarter that game? Maybe SF was just tired! Oh but also most of his worst runs also came in the 4th quarter. Maybe Rawls was tired and should have gone for more! The specific idea about "motivated" and "unmotivated" defenses is something I might try to tackle later, though. I think it's an interesting concept and I can't turn anything up on some preliminary googling! quote:Sorry if this has been asked before didn't see it, is there a way to possibly request a short summary of a player that played close to but a little over 6 years ago? Was gonna ask about Shaun Alexander Unfortunately, the database I'm working with only goes back to 2010. But I'd love to have data that goes back further! There's a lot of stuff that I've wanted to address, but didn't have the data to do so. But I'm happy to consider working with whatever data folks can get for me! If you just wanted to see the basics for Shaun Alexander, I'd just need play-by-play data for his career (something like year / week / down / distance / fieldposition / yards, for each attempt).
|
|
#
?
May 15, 2016 21:21
|
|
|
Yeah that's all totally fair, of course. I think the statistical analysis is useful regardless. I just thought it would be interesting to look into the situation, since I remembered that game. As an aside, in obscure RB news: australian Rugby star Jarryd Hayne is retiring from american Football (so he can go play rugby in the olympics). I'm disappointed because I really wanted him to earn a playing spot and put together enough data to be usefully analyzed. After playing a handful of games at the beginning of the season last year, Hayne had some fumbles and landed on SF's practice squad. I felt like the data set was too small to say for sure that he was going to have an ongoing fumble problem, and meanwhile, he showed a ton of promise with his unique skillset. SF didn't need him, though. I think if he'd landed on a more RB-needy team with better coaching, he might have had a real opportunity.
|
|
#
?
May 15, 2016 22:16
|
|
|
The 49ers were literally signing multiple RBs off the street while Hayne chilled on the practice squad. He styled on preseason roster fodder and got hearts aflutter but he was unlikely to be anything of value.
|
|
#
?
May 15, 2016 22:19
|
|
|
Yeah because he fumbled two or three times, at least one of which was just a strip that could have hit anyone, but one of which was a really big and public muffed punt/kick reception that was pretty embarrassing. Which is bad, but I don't think he had a decent chance to get comfortable and show his stuff. And the niners were (and perhaps still are) very very badly coached last year, and Trent Ballke who needs to die in a fire. None of which means Hayne should have been rostered and played more, of course. I'm just sad because I was rooting for him and he flashed some really interesting poo poo on the field in his very limited opportunities.
|
|
#
?
May 16, 2016 19:57
|
|
|
I mean he Willy Maysed a punt which might be the most impressive thing I've ever seen. I feel you in that respect, but he had a long long way to go to be anything.
|
|
#
?
May 16, 2016 20:21
|
|
|
The Outliers: The Greatest Single-Game Rushing Performances of 2014 Last time we developed a method for estimating the relative impossibility of top-tier single game rushing performances. Specifically, we derived the odds ratio between the game the player had and a game of “meh” runs. But measures like this always have a habit of introducing surprises once you actually start looking at the data. 2014 definitely had a pretty clear pattern of the top games. The top five are all high-volume games where a solid output was relentlessly maintained for an impossibly long time, and the next five are all explosive games where players go bananas with 20-30 carries. So rather than combine the “batting title” games and the “home run title” games, we’re going to spread the love and do two top-five lists in parallel. One for the volume games (with well over 30+ carries) and one for the flash games (with 20-30 carries). --- Volume Game #5 – DeMarco Murray @ CHI, week 14 Q: What do you get when you put the best offensive line in the league up against the worst defense in the league? A:  Murray had a ridiculous 41 touches this game (32 carries, nine receptions), gaining 228 total yards (179 on the ground). As far as Thursday football goes, this one was fun to watch. The Cowboys had so much swagger this game, in one 2nd-quarter drive when the game was still tied, Dallas had Murray go for it on a 4th and 1 from the CHI 13 (forgoing the easy field goal), and then four plays later, had Murray go for it again on another 4th and 1, this time from the goal line (which Murray converted for a touchdown). He also, bizarrely, had a run for a 14-yard loss.  This game allowed Murray to keep pace for his franchise-record rushing season, but this game was notable for Murray’s franchise record for number of 100-yard rushing games in a season, breaking Emmitt Smith’s old record of nine (this was Murray’s 10th). --- Flash Game #5 – Jeremy Hill @ CLE, week 15 One of my favorite ground games of the year. JFF makes his debut for Cleveland, completing a whopping 10 passes en route to a 30-0 blowout. And on the other side of the ball, Jeremy Hill quietly continued a special rookie campaign. Because Jeremy Hill was the best short-yardage runner of 2014. And it wasn’t close. Hill’s final stat line for this game was 25 carries for 148 yards and two touchdowns. Not bad for a rookie, right? Here’s the game log: -3, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 4, 5, 5, 5, 6, 6, 7, 7, 7, 8, 9, 16, 20, 21 On average in the NFL, about half of all carries are over by two yards. Of the 25 carries Hill took, all but one went 2 yards or more. Imagine flipping a coin 25 times and getting “heads” on 24 of them.  Only, this performance was even more improbable than that. Because he also broke out 5+ yard runs and 15+ yard runs at double the average rate. Only a handful of runs made the highlight reel, but this is a dominating performance from a back who refused to go down without gaining the extra yard. --- Volume Game #4 – C.J. Anderson @ KC, week 13 When the 8-3 Broncos went up against the 7-4 Chiefs on Sunday Night Football, Denver desperately needed their young running back corp to play big. The KC defense was playing some outstanding football, yielding the fewest passing yards in the league up until that point. Combined with plummeting temperatures, it was clear that this game would be won or lost around the run. Of course, Denver had just lost their #1 and #2 running backs to injury. That meant that the game plan was a) stop Jamaal Charles, who was averaging about a hundred yards and a touchdown per game at the time, and b) win on the ground with undrafted rookie Juwan Thompson and undrafted third-string Sophomore CJ Anderson. Charles ended the game with 10 carries for 35 yards. CJA finished with 168 yards on 32 carries. Though Anderson broke out six 10+ yard runs, his longest run of the day was only 20 yards. This was a game won on the back of relentless 3-8 yard gains.  Bizarro fact of the game: CJA took 3 carries from the DEN 41 in the 2nd quarter, spread across two different drives. They went for 0 yards, 0 yards, and 1 yard, the only 0-1 yard carries he had all game, and half of the carries he had that failed to gain at least 2 yards. That same quarter, Andre Caldwell drew a penalty from the DEN 41 for going out of bounds on a punt. The next play, depth safety and special-teams dude David Bruton took a direct snap on a fake punt and carried it for the first that evaded CJA from the spot all quarter. --- Flash Game #4 – C.J. Anderson @ MIA, week 12 Of course, perhaps we shouldn’t be too surprised by CJA’s big game against KC, considering the 195 scrimmage yards he dropped on the Dolphins the previous week. It was a barnstorming high-scoring game with a 4th quarter comeback and plenty of highlights, and CJA showed up in kind, posting three big runs of 20+ yards, a touchdown from 10 yards out, and a tense 4th down conversion at the end of the 3rd that kept Denver alive with an 8-point drive (dropping the deficit to 3).  CJA ended the game having burst into the secondary for 8+ yard runs at twice the average rate. --- Volume Game #3 – Rashad Jennings vs HOU, Week 3 I love to see scrappy old vets claw their way into grind-em-out wins like this one. Jennings dragged the Giants kicking and screaming to their first win of the season with 34 carries for 176 yards and a touchdown. It would be the only 100-yard rushing game Jennings would have all year. The way Jennings ground out respectable little runs was almost comical. In 34 attempts, only a single run went further than 15 yards. That would be the 18-yard run from the Giants own 21 that was immediately followed by 3 consecutive incomplete passes and a 5-yard penalty. It would be his longest run of the season. But to his credit, despite shouldering such a big workload, Jennings stayed a step ahead the entire game. It’s like he took the entire run distribution and shifted it a yard to the right:  About one third of Jennings’ runs first a first down for his entire 167-carry season occurred during this game. You go old man. --- Flash Game #3 – Le’Veon Bell @ CIN, week 14 gently caress yeah, now we’re talking.  For a three-game stretch in the middle of the 2014 season, Bell was untouchable. In consecutive games against TEN, NO, and CIN, Bell accrued a total of 711 yards from scrimmage and five touchdowns. This game was the capstone of that stretch. Bell took 26 carries for 185 yards. This is what that looked like:  Nearly all of this was gained while the game was still competitive. 20 of the 26 carries took place while the Steelers were behind, and 22 occurred when the game was within 1 touchdown. He also added six catches for 50 yards to that total. He ended the day with 235 combined yards and three touchdowns, and was named NFL Ground Player of the Week and AFC Offensive Player of the Week for his efforts. After this game, there are now two players in the history of the NFL who have broken 200 combined yards in three consecutive games: Walter Payton, and Le’Veon Bell. By the end of the regular season a few weeks later, Bell had broken the single-season franchise records for yards from scrimmage (2,215), first downs (114), receptions by a running back (83) and receiving yards by a running back (854). --- Volume Game #2 – Le’Veon Bell @ TEN, Week 11 Before the records started to fall, Bell also put in one of the gutsiest workhorse performances of the season. Through the first three quarters of play, Bell took 21 carries for 130 yards, and was busy juking dudes out of their shoes.  But a sputtering Steelers offense had left them behind 13-24. So Bell stepped it up to another gear with some 4th-quarter heroics. Here is the entire drive for the first touchdown of the 4th quarter: quote:• 1-10-PIT 28; 7-B.Roethlisberger pass incomplete deep left to 11-M.Wheaton. Bell personally accounts for 52 of 72 team yards and five of seven plays for gain, scoring the touchdown by the tips of his fingers in a desperate full extension.  Score: PIT 20 – 24 TEN. Steelers defense comes through with a three-and-out. That’s followed by the second drive of the 4th quarter: quote:• 1-10-PIT 35; 26-L.Bell right guard to PIT 45 for 10 yards (39-D.Stafford). Bell accounts for 38 of 65 team yards and five of seven plays for gain. Score: PIT 27 – 24. Steelers get the ball with 6:58 left on the game, and the Titans have two timeouts left. PIT needs to salt away as much of the remaining time as possible. An exhausted Bell can’t rest yet. It was the final drive of the game: quote:•1-10-PIT 19 (6:58) 26-L.Bell up the middle to PIT 25 for 6 yards (93-M.Martin). Bell finished the game with 33 carries for 204 yards and a touchdown.  --- Flash Game #2 – Mark Ingram vs GB, week 8 Eddie Lacy rumbled for 182 combined yards this game. The Packers got blown the gently caress up anyways, due in large part to Lacy’s former Alabama teammate. Ingram took the field for 54 of 68 offensive snaps, pulling 24 carries for 172 rushing yards and a touchdown. Two things are notable about Ingram’s performance this game. 1) An absurd number of runs for exactly 4 yard (one every four attempts), and 2) two and half times as many 9+ yard runs as you’d expect for this number of carries.  Fun fact: there have only been three regular season games in NFL history without a punt. The Packers participated in two of them within a month of one another. --- Volume Game #1 – Jonas Gray @ IND, Week 11 I mean really, could it have been anything else? Gray got his first NFL carry in week 7 against the Jets. 3 attempts for 12 yards. Four weeks later, Belichick rides him to a 38-carry, 201-yard, 4-touchdown game.  It was as many rushing touchdowns as the rest of the NFL combined that weekend. It doubled the Pats rushing TD total for the year.  The Pats had such a simple game plan: load the front with a sixth lineman and two blocking tight ends, then run loving power until the game was over. They had an extra lineman on 38 of 74 snaps. And the Colts just couldn’t stop it. All but 3 of his 38 carries gained yardage. 29 carries went for 3 yards or more. It was a complete and total humiliation. https://twitter.com/joshkatzowitz/status/534194636084891650/photo/1 Jonas Gray was inactive for two of the remaining regular-season games, didn’t play in the superbowl, then got cut from the team. --- Flash Game #1 – DeMarco Murray @ TEN, week 2 After losing their opener against San Francisco, the Cowboys took this bounceback as a statement of intent: this year, they’re going to run the danged ball. Murray was immense. Half of his carries went six yards or more. Read that again. Half.  He posted 167 yards on 29 carries (with 10 first downs), proving that he could carry the load without sacrificing his efficiency. Murray went on to string together a record-breaking six more 100-yard rushing games in a row (for a total of 8 to start the season) en route to finishing the season #1 in yards from scrimmage. This offseason, the Titans traded for Murray. --- Honorable Mentions: Best Game under 20 carries: Lamar Miller, week 7, narrowly beating DeMarco Murray, week 8. Best Game under 15 carries: LeGarrette Blount, week 3. Best Game under 10 carries: Shane Vereen, week 5 (game log: 0, 5, 7, 9, 10, 11, 14, 15, 19).  Best Game by a Player Over 30: Frank Gore, week 16 and week 17. Most Appearances in the top 20: DeMarco Murray with four, weeks 2, 4, 5, and 14. Worst Game of the Year: CJ Spiller, week 5 (game log: -2, 0, 1, 1, 1, 1, 1, 2, 2), and Trent Richardson, week 11 (game log: -3, -2, 0, 0, 1, 2, 2) Forever_Peace fucked around with this message at 04:54 on May 18, 2016 |
|
#
?
May 18, 2016 04:43
|
|
|
Can I get a drat trigger warning on the bell game ffs
|
|
#
?
May 18, 2016 04:51
|
|
|
Hell yeah Rawls, that was a great game last year. Hope he keeps it up. 
|
|
#
?
May 18, 2016 13:11
|
|
|
I noticed you did a Volume game and Flash game top 5 for 2014. Do you plan to revisit 2015 to do similar style breakdown of the top 5?
|
|
#
?
May 18, 2016 15:47
|
|
|
MacheteZombie posted:I noticed you did a Volume game and Flash game top 5 for 2014. Do you plan to revisit 2015 to do similar style breakdown of the top 5? Nah. I'm letting the data guide me to the right narratives. The distinction here was just really evident in 2014. It felt unfair to leave out the flash games. They were just so fun! By contrast, in 2015, only one of the top 5 overall games was over 30 carries. The top 5 of 2013 is also a pretty good mix. But the notable thing in 2013 was two really special games under 20 carries, so 2013 will have a top-5 plus the pair of special mentions. I'm thinking 2012 will probably be listed in reverse. Jamaal Charles had three games in the top 10, so I'm thinking it might be a good opportunity for a Charles love-fest to wrap things up. 2011 has a pretty clear top 5 without too many special things to note. 2010 will be tough. There's a clear top 3, then 6 similarly impressive games. I'll have to think on it.
|
|
#
?
May 18, 2016 16:05
|
|
|
Oh man, every single one of the top-5 2013 games were really, really fun games. 2013 was a great year for football. Having a good time with this one.
|
|
#
?
May 20, 2016 18:02
|
|
|
|
| # ? Apr 19, 2024 22:17 |
|
|
One post closer to next page, since this page is starting to lag for me... btw, DotY started https://forums.somethingawful.com/showthread.php?threadid=3776557
|
|
#
?
May 20, 2016 18:26
|
|