I can't fucking believe you're throwing a fit at FO for calling Morris the 5th best rusher in the NFL this year. That's basically the same as his Yards/Attempt rate him. What the fuck do you want?
Christ. I'm not throwing a fit. I'm curious as to why he's ranked overall the 7th best running back, behind Joique Bell. It makes no sense to me. I find FO utterly fascinating because they present these statistics, all wrapped up...that when you take a second to look behind the curtain, don't make a lot of sense in the end. And since they don't show us their formulas, none of it is transparent. Like how their ratings hold the 2000 Ravens defense as the 11th best defense since 1991. Hell, it's one of the best of all time, but not according to them. Why are they ranked lower than the 1991 Saints defense? Or the 2009 New York Jets defense? Or hell, the 2012 Chicago Bears defense! Does that make any sense whatsoever to you?
FO doesn't count lost fumbles, just fumbles (whether they're recovered or not). And looking at ESPN's stats, RG3 has 12 total fumbles. Luck had 10 total fumbles.
re Bradshaw/Morris/Bell: Morris: 255 Rushing DYAR, -1 Receiving DYAR, 254 total DYAR. Bradshaw: 230 Rushing DYAR, 35 Receiving DYAR, 265 total DYAR Bell: 71 Rushing DYAR, 193 Receiving DYAR, 264 total DYAR So basically: Morris is fantastic at rushing, better than either guy. He was useless as a receiver. Bell is OK not great at rushing but a fantastic receiver. DYAR measures total contribution; it turns out that Bell (and Bradshaw), by virtue of their ability to function as receivers, contributed more than Morris did. That's not to take away from Morris's (very impressive) rushing achievements! That's just to say that maybe he should catch balls more. :) Also: the simple fact is that - apparently - Washington faced an easier schedule in terms of rushing defenses. So Morris's 1500 yards simply count for less than Bradshaw's.
It's always good to know you're on board for an intelligent discussion, Bill! Note: I acknowledge I'm not sold that DVOA is valid for cross-season comparisons. :)
What the hell are you looking at? This is his FO stat line: Player Team DYAR Rk YAR Rk DVOA Rk VOA Runs Yards EYds TD FUM Suc Rate Rk A.Peterson MIN 459 1 359 1 24.9% 2 17.5% 348 2,097 1,923 12 3 49% 14 M.Lynch SEA 361 2 269 4 19.2% 3 12.0% 315 1,592 1,702 11 4 50% 11 C.J.Spiller BUF 301 3 292 2 27.5% 1 26.3% 207 1,249 1,205 6 2 55% 3 F.Gore SF 268 4 214 5 17.4% 4 12.0% 258 1,213 1,323 8 3 48% 17 A.Morris WAS 255 5 275 3 10.3% 8 11.7% 335 1,614 1,572 13 4 52% 7 He's 5th in defense adjusted above replacement, 3rd in raw, 8th in DVOA. He's 7th in success rate, which has far more to do with the line (Denver's backs are first and second.) Bell isn't even on the fucking list, he only has 82 carries.
That's actually a much tougher thing to quantify. FO has tried to come up with some stats to measure offensive/defensive lines (adjusted line yards), but they're a bit hazier and they aren't incorporated into DVOA/DYAR except insofar as the defensive line contributes to the defensive adjustment. A fantastic O-line will just show up as really good rushing DYAR/DVOA though.
Fair enough. Just bizarre that having actually only lost 2 of those fumbles, all of a sudden 600 yards and 2 more TDs is rated as less valuable. When you look at game impact, defenses don't gameplan for Luck's rushing...they do for RG3. But that's reality, I suppose? :)
Thread moving fast! Who was this in reply to? You're right in that Bell didn't even make 100 carries; his primary contribution was as a receiver.
1) FO doesn't count lost fumbles, just fumbles. Fumble recovery is a random event, if you fumble in the first place you failed. 2) Therefore it's a delta of seven fumbles, not two. 3) Defensive adjustments matter; I gather the Redskins played a weak schedule of rushing defenses.
Yeah...see I don't see how rushing for 1600 yards and 13 TDs counts for 255 "points", but receiving 52 passes for 485 yards and 0 receiving TDs counts as "193" points in those "Top Running Back" category...
That does certainly seem strange; I'll ask my friend if he can offer a quick accounting. My guess: opponent adjustments matter.
One suspicion: Bell's receiving stats are going to be rated relative to other running backs, not wide receivers. In that case, his lofty(ish) DYAR numbers for the receiving stats make sense (because they're relative to a replacement running back). It still definitely seems overstated though; I'm trying to get my friend to explain it better. He did confirm the Skins played a below average schedule of rushing defenses, which is why Morris and Griffin don't get as much credit DYAR-wise as their basic statlines indicate they might deserve.
What did you say about Morris on passing downs? That he wasn't even on the field? Why the fuck do you care about the combined rankings? Are you pissed that Sproles gets credit for being a receiver?
Well that's all you had to say! Incidentally I like FO for a lot of stats. I think DYAR is very useful.
I think it's kind of important to be really precise about how we're defining things. Which is tough when we get into terms like "better." In terms of running the ball, Morris was absolutely a beast this year. He had a fantastic rookie season . The reason he's not ranked as "better" than other guys in FO's rankings is basically: 1) Opponent adjustments. He ran up yards against soft rushing defenses. 2) He's basically replacement level as a receiver. #2 gets to be a bit controversial because obviously that's not his primary job. On the other hand, all things being equal, we'd rather the RB be a competent receiver than not. So while he doesn't get dinged for that (because DYAR is cumulative and you don't get dinged for things you don't do), he doesn't get credit for it either, while other guys who are good receivers get credit for it. Does that make them better? I duno. DYAR is an attempt to measure your total contribution over a season, according to DYAR Bradshaw contributed a bit more to his team than Morris did, adjusted for the defenses they played.
I'm sure they compare RBs to RBs, but that seems like a lot of "value" attached to what in the end is a pretty minimal number of yards and no touchdowns. Ask him if he'll share his formula...;) They did...though a lot of that is them putting up yards on every defense they faced...heck even against Tampa Bay (the #1 ranked run defense), the Skins put up 153 rushing yards when they met this year (I believe thats the most the Bucs gave up...even more than when they met Minnesota). By the way, thanks for engaging on this. And CB,when FO issues their "Top 5 RB" ranking, and ranks Joique Bell as a more valuable overall RB than Alfred Morris, I wonder why because that doesn't make much sense to me. I want to see why they value 485 receiving yards as they do. Its really less about Morris, and more about how do they reach their conclusions. Maybe you're comfortable not asking why, or you just don't care. If thats the case, ignore the conversation, step away, and let us discuss. I don't see why you're pitching a fit over my interest in learning more about their statistics.
Add in that Bradshaw played 2 less games than Morris, as well, which shouldn't affect DVOA, but apparently durability was one of the driving reasons behind DYAR...
Because you're acting entirely like the worst kind of NFL fan. Disingenuous about any kind of analysis that admits complexity rather than a top ten list. What part of overall value is so difficult for you to understand? Not interested in how a RB catches the ball? Then don't consult the combined values. If a guy is a third down back and is consistent in catching the ball for first downs, is that value or no?
You're completely missing the point here. Do I give a shit how some website ranks my RB? Or in the same vein if my QB wins Rookie of the year? No, I really couldn't care less because all I know is in the end I'm happy to have both on my team. What I'm interested in us seeing how FO actually is arriving at their values. Why are 485 receiving yards and no TDs worth 193 DYAR? How do they get to that number? How do they conclude that Brady had 6330 estimated yards? Where does that number come from? You may not care, and bully for you. But for me to understand if a statistic is actually valuable, you have to understand the formula and reasoning that creates it.
I've always been suspect of them, just as I'm suspect of ESPN's QBR rating. Simply because neither are transparent about the formulas driving their conclusions. It moved to absurd when I saw that FO rankings are citing three teams from this year as the greatest since 1991. And then when they ranked the 2000 Ravens defense as the 11th best defense of the past 20 years. Assertions like that are simply crazy. But that's what their statistics are telling them.
Re Morris/Bradshaw: here is my friend's short-version answer (he doens't have much time): Morris got stuffed a lot on third down, with a bunch of negative yards on third and short. Bradshaw had far fewer stuffs on third and short. Additionally, on third down Morris had no runs longer than seven yards. Bradshaw had five runs of 10 yards or more, including a 33 yarder. Factor in opponent adjustments and Bradshaw ends up being slightly "better." His subjective characterization of Morris on third down is "lousy." Converting downs matters a lot for FO's stats. re Luck/Griffin: it's entirely fumbles. Take away Griffin's 7 rushing fumbles and his rushing DYAR crushes Luck's. He hasn't gotten back to me on Bell, and I haven't asked about the cross-year comparisons; like I said I don't necessarily think DVOA supports them very well because a major factor is the current-season defensive adjustment.
Thanks. If its a really important metric, why aren't 3rd down conversions in the chart? Because all I can see is "Success percentage", which is more than just those 3rds...
Good question. Possibly because hscroll only allows so much. Also possibly: because they do have a premium product to sell. I think (though I'm not sure) that their premium product includes per-play stats, though I'm certain. At this point I'm speculating.
There is a lot of talk that Andy Reid will end up coaching the Cardinals, possibly with Tom Heckert (just fired by the Browns but previously with the Eagles) as GM. Reid to San Diego seems out despite his growing up in SoCal. The Chargers are an interesting team to watch. An article I read suggested that the Chargers may go for a big name this time around. While the Spanos family is still trying to keep the team in San Diego, they need some way to excite the fan base to get the support they need politically to get the stadium deal they want. And if they do end up moving to LA, they will need a coach who can get people excited there as well. So names like Chip Kelly and Jon Gruden seem like they might be associated with the job. On another note, the 2013 opponents are set (just not the dates). For the NFC West its a NFC South/AFC South year. So along with their regular division home and away games, the Seahawks face the Saints, Bucs, Vikings, Jags, and Titans at home. On the road they will visit the Giants, Panthers, Falcons, Texans, and Colts. They sure seem to play the Giants a lot. This will be third time in four years and fourth out of the last six there has been a Seahawks/Giants regular season game.
And they'll never shut up about it because Dilfer, Young, and some other ESPN NFL guys came up with it IIRC.
nah, nah, they didn't come up with it. But they are some fucking earnest cheerleaders for it. http://theclassical.org/articles/what-happened-to-footballs-next-great-stat Which is what infuriates me about it. Why are they cramming all of that into one rating? How are they assigning responsibility? It's a horrible mashup of decent ideas to make a useless mess. Which is part of the reason I'm being extra dickish to 8ball, because he's demanding a general overall ranking align with perceptions that are not at all general. When we think about great QBs, we think about specific things that they do well. Nobody should care about how Payton scrambles, so why the fuck are we figuring that into the statistic? Nobody cares what would happen if the 2000 Ravens walked out of the tunnel to play the 2012 Seahawks. It's stupid and pointless to cast that wide a net.
But that's exactly my point. Why is FO doing this? How do they assign values to such subjective measurements and come up with their rankings? If it's so "mathematical"...what's the goddamn formula for determining how valuable catching passes is versus running a ball for a running back? Remember, I'm not the one who is ranking teams and players...they are, and then they present their stats as "valid" and "determinative". Them asserting that the Seahawks are the 6th best team of the past 22 years makes me laugh, but also makes me wonder how valid their statistics are, because no one in their right mind would think this. Why should I put any more credence to their stats than I do to fat Pete Prisco's opinions...because as far as I can tell, both are equally shit?
It's starting to be hard to take your objections seriously when they seem to be kind of facile; it seems it boils down to: "I want to see the numbers" and "Seattle doesn't feel like the sixth best team of the past twenty-four years." But I'll try! With regards to the first: from a strictly academic standpoint I'd like to see the numbers as well! But I can also recognize that it's a proprietary system that's the basis of a good chunk of their revenue model. Their methodology is very much a matter of public record: figure out what "average" performance is in every situation and then compare a team/player's performance to that average to determine how much value they contribute. This is a tried-and-true method of sports performance analysis, pioneered (as far as I know) by the sabermetrics guys for baseball in the 1980s and adapted to football. You haven't really offered up any cogent criticism of this method of analysis, nor have you really demonstrated an awareness of how it works. Hence: getting hard to take you seriously. With regards to the latter: I get that you don't like the results it comes up with. That's not a reason to dismiss a model, it's a reason to explore why the model may have come up with those results. It's also a reason to question your own beliefs and assumptions. Your starting point is that the 2000 Ravens are definitely better than the 2012 Seahawks, but you don't have much to back that up other than you feel that way. That's perfectly fine, but you should acknowledge that you're not really even attempting any analysis here, you're just shooting from your gut in the same way that everyone does when they come up with various power-rankings style lists.
In other news, a player in the worst defense in the history of the universe thinks it's all Spagnulo's fault and wish Gregg Williams was back. I could not make this shit up if I tried. Since the Saints are in a bad cap space place (second after the Jets) I'd suggest it's time to wave goodbye to Vilma, Harper, Smith and any other defensive player who is making lots of money and start fresh.
