Om Posted October 22, 2005 Author Share Posted October 22, 2005 I don't think so. In my opinion, your use of the study correlating turnovers to wins marred an otherwise well-written and convincing argument. I was just offering a little constructive criticism.In order to measure a factor, it must first be isolated from other factors. In this case, we know that turnovers often lead to scores. We also know that high scoring teams get more turnovers simply by establishing leads in ballgames and forcing their opponents to take more risks in order to catch up. A valid study would find a way to isolate one factor or the other. In asking me to come up with statistical proof that high scoring teams eventually force their opponents into turnovers, you are attempting to shift the burden of proof. I've given you a reason to doubt the validity of the study. You have agreed my reasoning is logical.... I've made my point. I have zero problem with constructive criticism. I just like it more when it correctly targets something I've actually said. My use of the turnover differential statistics were not offered as absolute proof of anything. And I made it very clear that I understood ANY statistic can and should rightfully be viewed with the idea of context in mind. I also said--and have repeated now several times, as well as provided more statistical evidence to back the contention--that the indisputable fact is that teams that lose the turnover battle, for whatever reason, tend to lose. And that Washington is an anomaly in that they are NOT. You came in and attempted to parse the statistic, which is fine, but to my mind failed to acknowledge the context in which I was used them and was kind of tangential. You suggested that turnovers tend to come when a team has a lead, can get aggressive defensively and force more turnovers. I agreed that sounded reasonable, but that it didn't seem to correlate to the actual games the Redskins had been involved in this year ... and that I thought to more adequately make your case, at least to my mind, you'd need to come up with some evidence. I didn't attempt to parse the stats. I merely pointed out what I'd found to be a very compelling pattern in them to make a point. You took it farther and did attempt to parse the stats. The "burden of proof" should fall upon the person making the specific claim, no? Link to comment Share on other sites More sharing options...
Oldfan Posted October 22, 2005 Share Posted October 22, 2005 I didn't attempt to parse the stats. I merely pointed out what I'd found to be a very compelling pattern in them to make a point. You took it farther and did attempt to parse the stats. The "burden of proof" should fall upon the person making the specific claim, no? You presented the study of the correlation between turnovers and wins and losses as valid evidence supporting your premise that the turnover differential has the strongest correlation to wins and losses. I gave you a logical reason to question the basis of the study. In effect, we know the numbers you offered exaggerrated the importance of turnovers. What we don't know is by how much. Is Cincinnati +16 on the turnover differential because they blew out their opponents, or did they blow out their opponents because they are +16? The truth is probably somewhere between the two. You offered a premise based on statistical evidence. I'm not claiming anything here except that the study you offered can't logically be trusted to reflect the true correlation between turnovers and wins and losses. Link to comment Share on other sites More sharing options...
Om Posted October 22, 2005 Author Share Posted October 22, 2005 You presented the study of the correlation between turnovers and wins and losses as valid evidence supporting your premise that the turnover differential has the strongest correlation to wins and losses.I gave you a logical reason to question the basis of the study. In effect, we know the numbers you offered exaggerrated the importance of turnovers. What we don't know is by how much. Is Cincinnati +16 on the turnover differential because they blew out their opponents, or did they blow out their opponents because they are +16? The truth is probably somewhere between the two. You offered a premise based on statistical evidence. I'm not claiming anything here except that the study you offered can't logically be trusted to reflect the true correlation between turnovers and wins and losses. Not surprisingly, I disagree. To me, poor turnover differential stats correlate overwhelmingly to wins and losses, a thesis I contend is supported overwhelmingly by the statistics the NFL provides. Honestly, I have yet to see anything in your argument that comes close to refuting that fact. If your broader point is that statistics without context can be misleading, I agree entirely. But as your point seems to be that use of these particular statistics at all, as part of the much broader argument I have made, are without merit simply because some statistics can be misleading, I disagree entirely. Again, I have yet to deny that the reasons for turnover differential in and of themselves can be parsed. You, conversely (oops, there's that word again), have yet to acknowledge what appears to me a VERY strong correlation---regardless of the reasons for them---between turnover differential and wins and losses. At this point, I think (hope) I've made my point, so I'm prepared to let you have the last word on the matter if you choose. I can be cool that way sometimes. Link to comment Share on other sites More sharing options...
Oldfan Posted October 22, 2005 Share Posted October 22, 2005 At this point, I think (hope) I've made my point, so I'm prepared to let you have the last word on the matter if you choose. I'll gratefully accept the Last Word and freely acknowledge that it comes with no concession on your part. In order to measure something, it must be first isolated, when that cannot be done, when you are combining two or more factors in your measurement, the result is an unusable number. There's no question that turnovers are an important factor in determining the outcome of football games. The question is...how important are they? I think that, due largely to flawed studies like the one you offered, their importance is generally overrated. Link to comment Share on other sites More sharing options...
disneyskin Posted October 22, 2005 Share Posted October 22, 2005 Thinks Om, I finally got my fix and I feel much better now! :logo: Link to comment Share on other sites More sharing options...
beachwoodskinfan Posted October 22, 2005 Share Posted October 22, 2005 :eaglesuck :applause: :applause: :applause: :applause: :applause: :applause: :applause: :applause: :applause: :allhail: :49ersuck: :gaintsuck Dude, this is THE most well written and on the money article that I've read in YEARS. I'm not just saying that because you give hope to Redskin fans everywhere, If you're not already, you should seriously consider being a journalist or a sports columnist. " HAIL TO THE REDSKINS' Wash.- 42 S.F.- 6 Link to comment Share on other sites More sharing options...
BigSlamu Posted October 23, 2005 Share Posted October 23, 2005 Om - Not only do you make many excellent observations in this article you do a great job of communicating your reasoning behind them. This should not only be required reading for ES members but anyone who desires to be or calls themselves a sports journalist.:notworthy:applause::thumbsup: Very well done indeed. Link to comment Share on other sites More sharing options...
whitejimmy Posted October 23, 2005 Share Posted October 23, 2005 I'll gratefully accept the Last Word and freely acknowledge that it comes with no concession on your part.In order to measure something, it must be first isolated, when that cannot be done, when you are combining two or more factors in your measurement, the result is an unusable number. There's no question that turnovers are an important factor in determining the outcome of football games. The question is...how important are they? I think that, due largely to flawed studies like the one you offered, their importance is generally overrated. If I can jump in, I love a debate over logic. I believe Om's position that winning the turnover battle and winning games are strongly correlated is very true and has been statistically demonstrated time and time again by analysts. I think what Oldfan is getting at is the more important question: although the two are correlated, is the relationship causal. It's somewhat a matter of semantics, really. I think this has been pointed out, but it's the same situation with rushing. You always hear stuff like: "when the teams run the ball 40+ times, they win 95% of the time." Well, it's because they're winning that they can run more; it's not necessarily the running that's causing the winning. This logic phallacy is called "Non Causa Pro Causa." That said, I think TO's cause losing more than losing causes TO's. Link to comment Share on other sites More sharing options...
Oldfan Posted October 23, 2005 Share Posted October 23, 2005 If I can jump in, I love a debate over logic. I believe Om's position that winning the turnover battle and winning games are strongly correlated is very true and has been statistically demonstrated time and time again by analysts. I think what Oldfan is getting at is the more important question: although the two are correlated, is the relationship causal. It's somewhat a matter of semantics, really.I think this has been pointed out, but it's the same situation with rushing. You always hear stuff like: "when the teams run the ball 40+ times, they win 95% of the time." Well, it's because they're winning that they can run more; it's not necessarily the running that's causing the winning. This logic phallacy is called "Non Causa Pro Causa." That said, I think TO's cause losing more than losing causes TO's. Jimmy, I think it's more than semantics. There's no point in doing the statistical study if the result isn't going to accurately quantify the causal relationship. Working with your analogy: I'm certain there have been football coaches for decades who were misled by the baloney statistic into thinking that running the football 40+ times was the key to winning. An over-emphasis on turnovers can lead to equally dumb stategies. Brett Favre is living proof that turnovers on offense need to be weighed in one hand with production in the other. It's the same on defense. Your CB has four INTs for the year, but how many TDs and first downs did he give up by jumping routes and coming up empty? Link to comment Share on other sites More sharing options...
whitejimmy Posted October 23, 2005 Share Posted October 23, 2005 Jimmy, I think it's more than semantics. There's no point in doing the statistical study if the result isn't going to accurately quantify the causal relationship.Working with your analogy: I'm certain there have been football coaches for decades who were misled by the baloney statistic into thinking that running the football 40+ times was the key to winning. An over-emphasis on turnovers can lead to equally dumb stategies. Brett Favre is living proof that turnovers on offense need to be weighed in one hand with production in the other. It's the same on defense. Your CB has four INTs for the year, but how many TDs and first downs did he give up by jumping routes and coming up empty? Oldfan, I agree it's more than semantics. That's why I said it's "somewhat" a matter of semantics. You were discussing "correlations" vice "true correlations." I tried to clarify the discussion by introducing what I felt were clearer terms, "causal relationships" vice what you were calling "true correlations." I jumped in this debate because I understand your point, and logically you have a sound argument. I introduced the 40+ rushing analogy to demonstrate your point. But as you said, the "truth remains somewhere between the two." Om is right, as evidenced by Redskin games this year, turnovers have led to losing more than losing has led to turnovers. So the "somewhere" is with turnovers leading to losses far more than the occasional desperate hail mary padding the TO ratio. Yes, there are plenty of misleading statistics, like the CB that has a high interception count but gives up 4 TDs a game. But high turnovers are not a baloney statistic, even though they can be slightly inflated by desperate hail mary's and the such. They do lead to losing. In short, I think your point is valid but not enough to discount the significance of losing the TO battle. Link to comment Share on other sites More sharing options...
Skinsphoto Posted October 23, 2005 Share Posted October 23, 2005 Not surprisingly, I disagree. To me, poor turnover differential stats correlate overwhelmingly to wins and losses, a thesis I contend is supported overwhelmingly by the statistics the NFL provides. Honestly, I have yet to see anything in your argument that comes close to refuting that fact. If your broader point is that statistics without context can be misleading, I agree entirely. But as your point seems to be that use of these particular statistics at all, as part of the much broader argument I have made, are without merit simply because some statistics can be misleading, I disagree entirely. Again, I have yet to deny that the reasons for turnover differential in and of themselves can be parsed. You, conversely (oops, there's that word again), have yet to acknowledge what appears to me a VERY strong correlation---regardless of the reasons for them---between turnover differential and wins and losses. At this point, I think (hope) I've made my point, so I'm prepared to let you have the last word on the matter if you choose. I can be cool that way sometimes. Om, stop wasting your valuable time. This is the most stupid argument I've ever heard by someone who has some measure of education. Your point was very simple as well as obvious, thank you for drumming up the stats. Well done again. Obviously, who ever wins the turnover battle has the best chance to win the football game. I guess that's why Vermiel and Gibbs share the information. But anyone who follows football know this. Why this guys doesn't understand, no one could have a clue. Duh. Link to comment Share on other sites More sharing options...
Skinsphoto Posted October 23, 2005 Share Posted October 23, 2005 I'll gratefully accept the Last Word and freely acknowledge that it comes with no concession on your part.In order to measure something, it must be first isolated, when that cannot be done, when you are combining two or more factors in your measurement, the result is an unusable number. There's no question that turnovers are an important factor in determining the outcome of football games. The question is...how important are they? I think that, due largely to flawed studies like the one you offered, their importance is generally overrated. "There's no question that turnovers are an important factor in determining the outcome of football games. The question is...how important are they?" Answer: "important." It looks like you answered your own question. Perhaps I'm missing something. Did somebody make you angry today? You seem far to educated to proliferate this type of illogic. No disrespect intented, but come on. Link to comment Share on other sites More sharing options...
Oldfan Posted October 23, 2005 Share Posted October 23, 2005 It looks like you answered your own question. Perhaps I'm missing something. You are missing something. You're missing the point of the discussion. How important are turnovers? On a scale of ten, are they a 2, 5 , 7 or 9? The fact is that we don't know because there is no valid statistical evidence to tell us. I don't know what I wrote that caused you to think I am angry. I made a suggestion that provoked an intelligent debate: I thought Om's piece was a good, well-presented argument...marred by the use of a study which we can say by logical deduction...exaggerrates the importance of turnovers...by how much we don't know. Link to comment Share on other sites More sharing options...
HapHaszard Posted October 23, 2005 Share Posted October 23, 2005 Om your article is a perfect example of why I come here, excellent. Link to comment Share on other sites More sharing options...
Oldfan Posted October 23, 2005 Share Posted October 23, 2005 In short, I think your point is valid but not enough to discount the significance of losing the TO battle. Agreed. However, let me add another insight. In the Kansas City game, the football was knocked out of Brunell's hand in the red zone...turnover...cost us points. Why did it happen? Third and ten...defensive pressure. Why were we in third and ten? We can't rush the ball well in the red zone. My point: Teams that are consistently successful on first and second downs have fewer sacks and turnovers. Focusing on the turnover itself and blaming the player who turns it over is mostly non-productive. Link to comment Share on other sites More sharing options...
redtilldead Posted October 23, 2005 Share Posted October 23, 2005 I have been a Skins fan since 1970.....enjoyed the highs and dealt with the lows. Great article OM. It combines knowledge with confidence....going to use both when talking around the water cooler about my favorite subject. Long live the Washington Redskins! Link to comment Share on other sites More sharing options...
canesluver Posted November 9, 2005 Share Posted November 9, 2005 Chicago -- Redskins fumble four times, lose two. Bears fumble thrice, lose one. Dallas -- Redskins fumble once, lose it. Cowboys fumble three times, recover them all. Seattle -- Redskins do not fumble. Seahawks fumble once, recover it. Denver -- Redskins fumble once, lose it. Broncos fumble once, recover it. Kansas City -- Redskins fumble 3 times, lose them all. Chiefs fumble twice, recover them both. The math: Washington has fumbled 9 times, and lost 7. Their opponents have fumbled 9 times, and lost a total of 1. Can anyone update this with the Giants and Philly games thrown in?? Link to comment Share on other sites More sharing options...
Henry Posted November 9, 2005 Share Posted November 9, 2005 Redskin have fumbled 15 times, lost 11. That's a 73% rate. The average is 13.2 fumbles, 6.5 lost. That's a 49% rate. Our opponents have now fumbled the ball 16 times, and lost 2. I don't know what the average is, but I'm sure it aint that. Link to comment Share on other sites More sharing options...
bird_1972 Posted November 9, 2005 Share Posted November 9, 2005 Redskin have fumbled 15 times, lost 11. That's a 73% rate. The average is 13.2 fumbles, 6.5 lost. That's a 49% rate. Are there any practice drills they can run this week to start improving these numbers? Link to comment Share on other sites More sharing options...
canesluver Posted November 9, 2005 Share Posted November 9, 2005 Redskin have fumbled 15 times, lost 11. That's a 73% rate. The average is 13.2 fumbles, 6.5 lost. That's a 49% rate. And add 4 more fumbles by our opponents, and 0 recovered by us. That's now 13 opponents fumbles, with 1 Skins recovery. Pretty unbelievable!! Link to comment Share on other sites More sharing options...
Henry Posted November 9, 2005 Share Posted November 9, 2005 And add 4 more fumbles by our opponents, and 0 recovered by us. That's now 13 opponents fumbles, with 1 Skins recovery. Pretty unbelievable!! Actually, I just added that in my edit. Including the San Fran game, it's 7 more fumbles with 1 recovery, or 16 and 2 overall. Link to comment Share on other sites More sharing options...
canesluver Posted November 9, 2005 Share Posted November 9, 2005 Actually, I just added that in my edit. Including the San Fran game, it's 7 more fumbles with 1 recovery, or 16 and 2 overall. I did forget SF game. :doh: Link to comment Share on other sites More sharing options...
budski Posted November 9, 2005 Share Posted November 9, 2005 Maybe they need to bring in the guy that stopped Byner from fumbling so much when he came over from the Browns. Link to comment Share on other sites More sharing options...
canesluver Posted November 9, 2005 Share Posted November 9, 2005 Redskin have fumbled 15 times, lost 11. That's a 73% rate. The average is 13.2 fumbles, 6.5 lost. (That's a 49% rate.) Our opponents have now fumbled the ball 16 times, and lost 2. I don't know what the average is, but I'm sure it aint that. Wouldn't it be 12.5% with an average of 49% ? Link to comment Share on other sites More sharing options...
Henry Posted November 9, 2005 Share Posted November 9, 2005 Wouldn't it be 12.5% with an average of 49% ? Sure. Let's go with that. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.