FearLess Price Posted July 25, 2018 Posted July 25, 2018 1 hour ago, oldmanfan said: I have to have a fairly good understanding of basic stats for my job, have to look at data and review submitted articles all the time. And I know if I had to figure out a way to statistically evaluate different college QBs I'd have to go to a professional statistician who would need to formulate a comprehensive multivariate analysis to cone up with any reasonable comparative data. If you read the article this is pretty much what beane says they did. They looked at all his passes and made their own stats according to what was happening in the play. They also mentioned that Wyoming doesnt incorporate many easy throws like other college systems. 1 hour ago, Hapless Bills Fan said: To be fair, groups like PFF and Football Outsiders (as I understand it) do employ professional statisticians and statistics like DVOA and DYAR are their attempt to formulate a comprehensive multivariate analysis. The problem is, for college football, there's far more variability in what types of offense colleges run and what quality of players they recruit. No one (as far as I know) has found a way to incorporate those variables into the equation. See above. It says in the article they made their own stats based off what they were seeing on field compared to the scheme and plays being run. Im still not souped on the Allen pick but it is what it is at this point. Wish em the best. I hope he wins the job in camp in a squash match, not just handed it. That would be a good sign.
jrober38 Posted July 25, 2018 Posted July 25, 2018 9 minutes ago, FeelingOnYouboty said: I was probably the biggest Josh Allen hater during the process but I'm a sucker and I've bought in. You dance with the one who brought ya. I want him to succeed not just because he's our guy now but man I want the media who've gone overboard with their analysis to really eat it. Some of the analysis is bordering on mean lol Plus like most NFL execs I'm a size queen. Just look at Josh Allen. That's 6'5 235 pounds of California beef. The boy can fill out a uniform and he has the best arm in the NFL. Cool number. I'm all in man. Hope he's good. I can't do it. I bought in on EJ even though the scouting reports said he was likely going to be crap. There's way too much that says Allen is going to be awful for me to get on board with another guy who needs to work on the basic fundamentals of throwing a football.
Sky Diver Posted July 25, 2018 Posted July 25, 2018 10 minutes ago, Bangarang said: Obviously you have concluded that Allen will be a superstar so you are going to reject anything that may say otherwise. I wonder what your opinion of Allen would be if the Jets drafted him instead.. I don’t know if he will be a superstar or not, but I am not going to rule it out because of QBASE.
oldmanfan Posted July 25, 2018 Posted July 25, 2018 8 minutes ago, FearLess Price said: If you read the article this is pretty much what beane says they did. They looked at all his passes and made their own stats according to what was happening in the play. They also mentioned that Wyoming doesnt incorporate many easy throws like other college systems. See above. It says in the article they made their own stats based off what they were seeing on field compared to the scheme and plays being run. Im still not souped on the Allen pick but it is what it is at this point. Wish em the best. I hope he wins the job in camp in a squash match, not just handed it. That would be a good sign. It would be interesting to talk to Beane and ask him what traits he values in his analysis and compare it to some of these sites.
TheElectricCompany Posted July 25, 2018 Posted July 25, 2018 7 minutes ago, Sky Diver said: They haven’t changed their model in 30 years? I find that hard to believe. You obviously put more faith in QBASE than I do. Fortunately for the Eagles they took QBASE with a grain of salt. QBASE is just one of many models that exist.I wonder how Wentz looked in others. If QBASE was an outlier, i think you can live with that. However, if every aggregate stat you find tells you the same thing, which in Allen's case, is that his odds of becoming elite are extremely low, that's a big problem.
Bangarang Posted July 25, 2018 Posted July 25, 2018 12 minutes ago, Sky Diver said: I don’t know if he will be a superstar or not, but I am not going to rule it out because of QBASE. Well in a previous post you said “when he becomes a superstar”..
Sky Diver Posted July 25, 2018 Posted July 25, 2018 12 minutes ago, TheElectricCompany said: QBASE is just one of many models that exist.I wonder how Wentz looked in others. If QBASE was an outlier, i think you can live with that. However, if every aggregate stat you find tells you the same thing, which in Allen's case, is that his odds of becoming elite are extremely low, that's a big problem. I doubt their model has considered every possible factor. If Allen becomes a superstar, which no one can rule out, then their model is flawed and they will have to modify it. Obviously QBASE missed the mark badly on Wentz. If they are clever, they are continuously fine tuning. 1
TheElectricCompany Posted July 25, 2018 Posted July 25, 2018 3 minutes ago, Sky Diver said: I doubt their model has considered every possible factor. If Allen becomes a superstar, which no one can rule out, then their model is flawed and they will have to modify it. Obviously QBASE missed the mark badly on Wentz. If they are clever, they are continuously fine tuning. I'm sure it hasn't. They have selected the stats they feel are predicative of future success. The model wouldn't necessarily be flawed if Allen succeeds. It's already been discussed ad nauseam that he if does so, he would be a big time outlier.
Sky Diver Posted July 25, 2018 Posted July 25, 2018 3 minutes ago, TheElectricCompany said: I'm sure it hasn't. They have selected the stats they feel are predicative of future success. The model wouldn't necessarily be flawed if Allen succeeds. It's already been discussed ad nauseam that he if does so, he would be a big time outlier. If he succeeds, their model is flawed and there is a factor or weighting that they missed. The blind faith in QBASE is truly astounding.
Just Joshin' Posted July 25, 2018 Posted July 25, 2018 1 hour ago, BillnutinHouston said: Im no analytics expert, but I think where analytics worshippers go too far is their belief that data from the past is always determinative going forward. It simply isn't, and the embodiment of this is the many late round picks (all of whom were measured analytically) who go on to outperform their higher-drafted peers at the NFL level. Its human nature to develop predictive models, and while these will inevitably get better with time, they will always be confounded by the unpredictability of the human element. The problem with models is that they include the persons bias in the model. If they think completion percentage is a better indicator than yards per attempt, the model will reflect that bias. A good model minimizes those biases and performs a validation to confirm previous results produce similar results. People tend to think a model must be good because it is math but not always true.
TheElectricCompany Posted July 25, 2018 Posted July 25, 2018 3 minutes ago, Sky Diver said: If he succeeds, their model is flawed and there is a factor or weighting that they missed. The blind faith in QBASE is truly astounding. If he succeeds, it will not be because there was a NCAA statistic that was missed. What would they put into their formula instead? Performance in games with wind gusts above 35mph (I don't know if he was any good there)? He's already had every measurable statistic dissected by the masses. As mentioned before, QBASE is one of many, and every aggregate stat I've seen comes to the same conclusion on Josh Allen.
Sky Diver Posted July 25, 2018 Posted July 25, 2018 (edited) 31 minutes ago, TheElectricCompany said: If he succeeds, it will not be because there was a NCAA statistic that was missed. What would they put into their formula instead? Performance in games with wind gusts above 35mph (I don't know if he was any good there)? He's already had every measurable statistic dissected by the masses. As mentioned before, QBASE is one of many, and every aggregate stat I've seen comes to the same conclusion on Josh Allen. I don't know what their model is or what factors they might be missing. As far as weather, it's certainly a factor. It's easier to throw a football in Hawaii than Alaska from Sept to December, for example. Wentz played in ND. Pretty nasty there in winter. Maybe they aren't weighting weather enough or it's not even in their model? Edited July 25, 2018 by Sky Diver
Bills Pimpin' Posted July 25, 2018 Posted July 25, 2018 It's always more accurate to say a player is going to be a bust since 80% of them are (QB's are prolly closer to 90%). That's why most media D-bags and forum trolls claim bust. It is not because they are really smart or did any work to come up with their opinion. They are the bunch that cares only about claiming "I told ya so". In fact, in this specific case, even if Josh Allen is a serviceable QB for 10 years they will say " I told ya he wouldn't be a hall of famer" or "I told ya (QB X) would be better" or "I told ya we shouldn't have picked him 7th". It's really a miserable existence for them and especially kool aid drinking homers Homers aren't all that great either (I tend to be one, I can't help it. I talked myself into Tyrod Taylor as a franchise QB and now I get to tell another fanbase he sucks), but it is a much better daily existence for sure. 1
bobblehead Posted July 25, 2018 Posted July 25, 2018 (edited) The QBASE data is concerning and the lack of quality competition needs to be considered too, but when people use EJ as a comparison that just shuts me down right there. Josh Allen is nothing like EJ Manuel. Allen’s competitiveness, confidence, history of carrying a team and his throwing ability is miles above what EJ ever had. Josh Allen doesn’t step forward on every throw and he’s like the obnoxious 11 year old who throws fastballs at you when you’re 5 feet away. That’s the deal-can he fix that? Edited July 25, 2018 by bobblehead
TheElectricCompany Posted July 25, 2018 Posted July 25, 2018 26 minutes ago, Sky Diver said: I don't know what their model is or what factors they might be missing. As far as weather, it's certainly a factor. It's easier to throw a football in Hawaii than Alaska from Sept to December, for example. Wentz played in ND. Pretty nasty there in winter. Maybe they aren't weighting weather enough or it's not even in their model? Who's playing in Alaska? Haha. I doubt performance in severe weather is a factor. However, if a prospect played in severe weather significantly more than his peers, I would take that into account. Allen's big snow game from this year Date Opponent Surface Result Att Comp Pct. Yards Yards/Att TD Int Rating 11/04/17 Colorado St. Turf W 16-13 20 10 50.0 138 6.9 0 0 107.96
Sky Diver Posted July 25, 2018 Posted July 25, 2018 (edited) QBASE gave Watson a 56.5% chance of being a bust. He looks pretty good this year before he was injured. QBASE makes adjustments for the quality of the offense the QB plays for the the quality of the defenses he plays against. Maybe QBASE underestimated the quality of the defenses and overestimated the quality of Allen's offensive teammates. Who knows. It's not an exact science. 9 minutes ago, TheElectricCompany said: Who's playing in Alaska? Haha. I doubt performance in severe weather is a factor. However, if a prospect played in severe weather significantly more than his peers, I would take that into account. Allen's big snow game from this year Date Opponent Surface Result Att Comp Pct. Yards Yards/Att TD Int Rating 11/04/17 Colorado St. Turf W 16-13 20 10 50.0 138 6.9 0 0 107.96 I'm sure that someone plays football in Alaska. Snow isn't the only factor. You are aware of the wind factor in Buffalo, right? I think that fair to say that it's easier to play in LA than Wyoming. Edited July 25, 2018 by Sky Diver
TheElectricCompany Posted July 25, 2018 Posted July 25, 2018 (edited) 10 minutes ago, Sky Diver said: QBASE gave Watson a 56.5% chance of being a bust. He looks pretty good this year before he was injured. QBASE makes adjustments for the quality of the offense the QB plays for the the quality of the defenses he plays against. Maybe QBASE underestimated the quality of the defenses and overestimated the quality of Allen's offensive teammates. Who knows. It's not an exact science. Weather isn't a factor in the performance of players and outcome of games? You can't be serious. You are aware of the wind factor in Buffalo, right? That has nothing to do with snow. Pump the brakes. Weather is absolutely a factor, but I don't think "performance in severe weather" is noteworthy when evaluating a NCAA player's odds of become an elite QB. I doubt that any of these statistical models added "performance in severe weather" into their equation. Edited July 25, 2018 by TheElectricCompany
Sky Diver Posted July 25, 2018 Posted July 25, 2018 Just now, TheElectricCompany said: Of course it's factor, although I would argue a pretty minor one. I doubt that any of these statistical models added "performance in severe weather" into their equation. Well, then the model is flawed. Maybe that's why it missed on Wentz?
billsintaiwan Posted July 25, 2018 Posted July 25, 2018 The Minnesota Twins analytics department had Nishioka as a can't miss prospect. I scouted him and wondered what the !@#$ they were thinking. Analytics are dangerous when used by people without a scouting background. They can be very useful, obviously.
Bangarang Posted July 25, 2018 Posted July 25, 2018 (edited) 8 minutes ago, Sky Diver said: Well, then the model is flawed. Maybe that's why it missed on Wentz? Stop trying to say it’s flawed because it may not be right 100% of the time. If stats were a guaranteed method of predicting an outcome then scouts would be taken off the road immediately. They are simply another tool for evaluation. Would you be singing the same tune if the numbers lined up with what you want to believe? Edited July 25, 2018 by Bangarang
Recommended Posts