Why the Giants have virtually no shot to win the Super Bowl

[SunnyDenmark]

I laud your effort, and I'm not going to look down on anyone who wants to look deeper into the game using stats. If you'll read some of the responses, I think you'll see why it's really not a useful stat for this particular game, even if it is somewhat interesting as a historical trend.

I applaude your skepticism as statistics can be manipulated to show just about anything. But I think you are being too critical with this one. Momentum is crucial in sports - agreed. But as Metro stated - the 'hot' teams would not be in the SB without being hot. Maybe you win one playoff game with a lucky bounce or bad penalty (vs opposition) or such a fluke. But not going to happen down the stretch in the season (to make the playoffs) AND 2-3 games in the playoffs to make the SB.

Moreover, you want to more heavily weigh the 2nd half of the season for these stats - due to the 'hot' theory. But there are MANY factors that affect the 2nd half of the season - (and not the first half) for good or for bad. For example warm weather teams having to play up north in the December cold, , domed teams who play indoors, the playoff contenders (the 'hot' teams playing down the stretch) playing against demorilized teams who have no shot at the postseason and therefore just 'mail it in', etc.

In statistics it is much better to use larger sample sizes (ie. whole season) as well as , as many seasons as possible (as in the 0 -37 statistic). So that is a pretty good sample size considering..

If one of the SB teams is in the top 10 (or better yet) top 5 in either offense or defense, that must mean they have a superior advantage that is very difficult to stop. Think shuting down Moss and Welker but then how can you stop them PLUS Mahroney AND Watson AND Faulk AND Stallworth, etc. If all the other 16 games it was hard to stop the offense / (or defense) and therefore gives it a top 5 or 10 ranking due to various mismatches - then it will be equally hard for the SB opponent to stop as well - no matter how 'hot' they are.

food for thought

PS. the most important argument I could see for weighing the most recent games (or 2nd half of the season) - would be if a significant injury occured to a player who was instrumental in creating the 'top 10' offensive or defensive stats....Thankfully for the Pats -this is not the case (TB will play and play well - sprained ankle or not). He doesn't have back to back bad games....

 

[Watson's IQ]

Of course not!

There is no guarantee that the sun won't go red giant tomorrow. There is no guarantee that the Earth won't be hit by a giant meteor next week. There is no guarantee that we won't all become zombies in an hour.

And there is certainly no guarantee as to who wins a football game.

The OP may have gone a little overboard in his title, but that doesn't change the fact that it hasn't happened once in 37 seasons.

Post hoc ergo proctor hoc. A team that isn't in the top 10 in scoring offense or scoring defense has never won the Super Bowl, therefore it must be true that this is the reason they haven't won. Right? If so, then this statistic definitely has some value.

Correlation does not imply causality is the stalwart saying of statisticians because it's true. Just because these two things are related doesn't mean that total scoring offense or defense is a particularly dominant cause of these teams not winning the Super Bowl.

People can say all day "well it's never happened", but I'm sorry, that does not make the statistic any more relevant than it is.

 

[PATRIOTSFANINPA]

Like my title says the Giants have VIRTUALLY no shot to win the Super Bowl,Thats means little based on my statistical historical Super Bowl Champions for the past 37 years and its something I consider a very strong stat - I am not saying they have NO chance because anything can happen but the Giants have to be the first ever since 1970 to do so by being ranked out of the top 10 in offense and defense and I don't think they have a good enough team to be the team to break the 37 year streak,Thats IMO and any argument won't change it but to discuss it from pro to con is still cool for those who think this stat is BS.

Las Vegas uses stats all the time in betting parlors so I think even though its a stat it can mean something if it keeps up for a long period of time,once again its IMO

I have made lots of money gambling on stats that appear strong in the past and in my many trips to Vegas and other gambling outlets and I intend to continue to use strong stats to help fuel my winning when it comes to sports betting and trying to figure out the outcome of games

For those of you who say stats are pathetic things and are of the past and mean nothing, does that mean that any Horse Race Handicapper who buys a racebook to check the previous races of a horse is an idiot? - there is no difference between stats like this and previous races of a racehorse which a handicapper uses a program to decide who is best to bet on judging mainly by his previous races.

 

[PatsFanSince74]

Regarding the "momentum" or "performance in the latter part of the season" point:

I went to Football Outsiders and, much to my surprise, they report that, according to their analytics, the Patriots actually outperformed the Giants in the last four games of the season. I went there expecting to find the opposite in order to refute the thread's hypothesis, so maybe I am wrong and the poster's point holds up.

 

[Watson's IQ]

I applaude your skepticism as statistics can be manipulated to show just about anything. But I think you are being too critical with this one. Momentum is crucial in sports - agreed. But as Metro stated - the 'hot' teams would not be in the SB without being hot. Maybe you win one playoff game with a lucky bounce or bad penalty (vs opposition) or such a fluke. But not going to happen down the stretch in the season (to make the playoffs) AND 2-3 games in the playoffs to make the SB.

Moreover, you want to more heavily weigh the 2nd half of the season for these stats - due to the 'hot' theory. But there are MANY factors that affect the 2nd half of the season - (and not the first half) for good or for bad. For example warm weather teams having to play up north in the December cold, , domed teams who play indoors, the playoff contenders (the 'hot' teams playing down the stretch) playing against demorilized teams who have no shot at the postseason and therefore just 'mail it in', etc.

In statistics it is much better to use larger sample sizes (ie. whole season) as well as , as many seasons as possible (as in the 0 -37 statistic). So that is a pretty good sample size considering..

If one of the SB teams is in the top 10 (or better yet) top 5 in either offense or defense, that must mean they have a superior advantage that is very difficult to stop. Think shuting down Moss and Welker but then how can you stop them PLUS Mahroney AND Watson AND Faulk AND Stallworth, etc. If all the other 16 games it was hard to stop the offense / (or defense) and therefore gives it a top 5 or 10 ranking due to various mismatches - then it will be equally hard for the SB opponent to stop as well - no matter how 'hot' they are.

food for thought

PS. the most important argument I could see for weighing the most recent games (or 2nd half of the season) - would be if a significant injury occured to a player who was instrumental in creating the 'top 10' offensive or defensive stats....Thankfully for the Pats -this is not the case (TB will play and play well - sprained ankle or not). He doesn't have back to back bad games....

Fair enough. I think we're going to have to agree to disagree on this one. You're making a correct point when you say that larger sample sizes are preferred, but I'm not sure it's applicable when the variables are constantly changing, as they inherently do over a course of a football season. I agree with you that an injury or something along those lines would argue more heavily for back-weighting the stats, but the lack of one doesn't remove the fact entirely either.

What would you say to my previous assertion that for this type of statistic, scoring offense and defense are relatively simplistic measures if we really want to evaluate the quality of the teams?

I would assert that even if we're not weighing the stats to compensate for the "hot factor", using some kind of offensive efficiency stat (yards per point, or something like that) would be far more demonstrative of the quality of the teams as opposed to "points scored" and "points allowed."

 

[PonyExpress]

Unless someone can give a viable, scientific explanation for that, I fail to see any "advantage" gained and consider it to be a statistical anomaly based on a small sample size.

Sorry, BF, I didn't mean to make you nervous.

 

[Watson's IQ]

Of course not!

There is no guarantee that the sun won't go red giant tomorrow. There is no guarantee that the Earth won't be hit by a giant meteor next week. There is no guarantee that we won't all become zombies in an hour.

And there is certainly no guarantee as to who wins a football game.

The OP may have gone a little overboard in his title, but that doesn't change the fact that it hasn't happened once in 41 seasons.

The last thing I'm doing is making an "anything can happen" argument, and if you think I was you should go back and re-read my posts.

What I'm saying is that if there's a statistic which argues for the point the OP is trying to make, it's not going to be this one.

When we go and look at more advanced statistics, they're STILL going to say the Patriots are significantly better, as we'd all expect. Look at Football Outsider's weighted DVOA for one example. My discussion relates to the statistics we're using as a basis for this discussion, which I don't think illustrate as much as the thread implies.

 

[PATRIOTSFANINPA]

I respect anyone who does not agree that an 0-37 stat is incredible but if the Giants win I will come back to swallow my stat and eat crow even though I am a Pats fan who would not be the only guy to eat crow but I don't think we have to worry about that happening because the Pats will win by AT LEAST 20 and I rarily ever predict a Patriots blowout by this much.

0 for 38 on the way baby!! :rocker:

 

[Oswlek]

Post hoc ergo proctor hoc. A team that isn't in the top 10 in scoring offense or scoring defense has never won the Super Bowl, therefore it must be true that this is the reason they haven't won. Right? If so, then this statistic definitely has some value.

Correlation does not imply causality is the stalwart saying of statisticians because it's true. Just because these two things are related doesn't mean that total scoring offense or defense is a particularly dominant cause of these teams not winning the Super Bowl.

People can say all day "well it's never happened", but I'm sorry, that does not make the statistic any more relevant than it is.

What are you talking about?

I specifically stated in several earlier posts that this was a non-causal stat.

 

[Watson's IQ]

I respect anyone who does not agree that an 0-37 stat is incredible but if the Giants win I will come back to swallow my stat and eat crow even though I am a Pats fan who would not be the only guy to eat crow but I don't think we have to worry about that happening because the Pats will win by AT LEAST 20 and I rarily ever predict a Patriots blowout by this much.

0 for 38 on the way baby!! :rocker:

Thanks, I certainly respect your efforts as well. Anyone who brings something more to the table than "PATS ARE BETTER!!! TEAM OF DESTINY!!" gets a gold star in my book.

That said, I always enjoy a little healthy debate on things like this, and I'm glad this has turned into a meeting of the minds of sorts. It's refreshing.

 

[Oswlek]

The last thing I'm doing is making an "anything can happen" argument, and if you think I was you should go back and re-read my posts.

What I'm saying is that if there's a statistic which argues for the point the OP is trying to make, it's not going to be this one.

When we go and look at more advanced statistics, they're STILL going to say the Patriots are significantly better, as we'd all expect. Look at Football Outsider's weighted DVOA for one example. My discussion relates to the statistics we're using as a basis for this discussion, which I don't think illustrate as much as the thread implies.

I understand exactly what you are saying. But you are trying to discredit what the OP said by making it seem as if this situation is unique in some way, which it isn't. EVERY team that lost the SB went in on a roll. EVERY team in this stat is ranked by season stats.

I understand that there is no causality, but that doesn't mean that a 37-0 correlation can be wholly disregarded as well.

 

[Watson's IQ]

What are you talking about?

I specifically stated in several earlier posts that this was a non-causal stat.

If you did say that, I apologize if I missed it. I saw you mention that in reference to the QB passing leader thing.

Well if we concede it's not causal, than while it's certainly a comforting stat for us Pats fans, it doesn't really give us any ammunition for why the Patriots should beat the Giants, from an analytical perspective.

If I have time later on I'm going to take some other stats and apply them to the framework he established in the thread, and I fully expect we'll see very similar results.

 

[BrianPat]

I believe the Pats will win the SB by a wide margin. However, when 2 teams face each other in the regular season in a game decided by 7 points or less, and then have a rematch in the SB, the team that lost the regular season game has won the SB rematch 6 out of 7 times. The only exception is the 1986 Broncos, who lost to the NYG by 3 and then lost the Sb 39-20, despite leading at halftime.

IOW, the NYG gain an advantage by having played the Pats close in the regular season. That advantage mitigates against any other statistical "rules" governing winning and losing the SB.

Disagree, the Giants weren't playing a complete Patriots team and I think the defense the Giants offense will see will be a defense it's never seen before.

 

[Watson's IQ]

I understand exactly what you are saying. But you are trying to discredit what the OP said by making it seem as if this situation is unique in some way, which it isn't. EVERY team that lost the SB went in on a roll. EVERY team in this stat is ranked by season stats.

I understand that there is no causality, but that doesn't mean that a 37-0 correlation can be wholly disregarded as well.

And I'm not wholly disregarding it, I just think that scoring offense and defense is not the best measure to use. I'm not doubting his math or that the statistical logic behind what he's saying is sound, I'm just not convinced yet about how much of a factor it is compared to others.

 

[Metaphors]

All of this discussion ignores special teams. There is a reasonable chance this game would be "one for the thumb" if not for Desmond Freaking Howard. Take away Hickson's return TD and the last game might have been very different. Don't ignore how the return game can be a major equalizer.

Also, the week 17 game was a 3 point game only because the Pats were willing to trade points for time. Why do you think Belichick was so angry with Rodney's stupid penalty (15 huge yards that took the Giants from their 39 to the NE 46 with no time off the clock). From that point on, it took the Giants 10 plays, 3 minutes and one timeout to go the rest of the way. Lost the battle and won the war.

The Giants were at home, had a QB with a 118 rating, went 6-10 on 3rd down, went 4-4 in the red zone, had a return TD...and lost the game! The Giants made big plays to keep it close, but that doesn't necessarily translate well to the next game. The factors mentioned above will likely not turn up in the SB:

• Not at home (don't give me the "road-field advantage" when you get close road wins against WAS, MIA, CHI, PHI and DET)
• Eli's rating was 74 (99 post-epiphany)
• 3rd down conversion rate was 42% (41% after the awakening)
• Red zone conversion rate was less than 100% (don't have stats handy)
• Return TDs other than Hickson's...bubkas

So can the Giants beat the Pats in the SB? Sure. Can they replicate week 17? Almost certainly not. The only "blueprint" week 17 gave the Giants was they need to get multiple big plays to hang with the Pats. Now they just need to find some a week from Sunday.

 

[Oswlek]

All of this discussion ignores special teams. There is a reasonable chance this game would be "one for the thumb" if not for Desmond Freaking Howard. Take away Hickson's return TD and the last game might have been very different. Don't ignore how the return game can be a major equalizer.

Also, the week 17 game was a 3 point game only because the Pats were willing to trade points for time. Why do you think Belichick was so angry with Rodney's stupid penalty (15 huge yards that took the Giants from their 39 to the NE 46 with no time off the clock). From that point on, it took the Giants 10 plays, 3 minutes and one timeout to go the rest of the way. Lost the battle and won the war.

The Giants were at home, had a QB with a 118 rating, went 6-10 on 3rd down, went 4-4 in the red zone, had a return TD...and lost the game! The Giants made big plays to keep it close, but that doesn't necessarily translate well to the next game. The factors mentioned above will likely not turn up in the SB:

• Not at home (don't give me the "road-field advantage" when you get close road wins against WAS, MIA, CHI, PHI and DET)
• Eli's rating was 74 (99 post-epiphany)
• 3rd down conversion rate was 42% (41% after the awakening)
• Red zone conversion rate was less than 100% (don't have stats handy)
• Return TDs other than Hickson's...bubkas

So can the Giants beat the Pats in the SB? Sure. Can they replicate week 17? Almost certainly not. The only "blueprint" week 17 gave the Giants was they need to get multiple big plays to hang with the Pats. Now they just need to find some a week from Sunday.

When using points scored, special teams are factored in by default.

 

[Watson's IQ]

When using points scored, special teams are factored in by default.

Yes, and they're credited to offensive points scored and defensive points allowed, even though those units had nothing to do with the score. Similarly, a defensive touchdown is factored into "points scored" even though the defense got it. That's a factor in why most use yards gained and allowed per game as the benchmark in team rankings.

I'm not saying at all that these stats don't speak to the quality of the team, they do, just that they're not always the best measure for predicting future success.

 

[SunnyDenmark]

What would you say to my previous assertion that for this type of statistic, scoring offense and defense are relatively simplistic measures if we really want to evaluate the quality of the teams?

I would assert that even if we're not weighing the stats to compensate for the "hot factor", using some kind of offensive efficiency stat (yards per point, or something like that) would be far more demonstrative of the quality of the teams as opposed to "points scored" and "points allowed."

Sorry but I cannot honestly reply. One hears scoring offense and defense bantered around alot in the press. I dont know if it is more valid than other metrics. My gut feeling is that it is more valid than yards gained and yards given up as it is the final score that gives the 'W'. But as you mentioned in another post, Football Outsiders uses some pretty advanced stats that probably could tell more. Unfortunately, those metrics are relatively new and therefore haven't been used for 37 seasons. So unless someone wants to go back all 37 seasons and apply the new statistics , we are left using the 'old traditional' ones. And that 0-37 sounds good to me!

You know what also sounds good to me: 19-0. Yeah Baby!

 

[Oswlek]

Yes, and they're credited to offensive points scored and defensive points allowed, even though those units had nothing to do with the score. Similarly, a defensive touchdown is factored into "points scored" even though the defense got it. That's a factor in why most use yards gained and allowed per game as the benchmark in team rankings.

I'm not saying at all that these stats don't speak to the quality of the team, they do, just that they're not always the best measure for predicting future success.

I disagree somewhat. I agree that STs (along with offensive TOs and various other factors) will influence points score and thus, ranking units based on points allowed/scored is falwed.

However, if you are looking at just two teams, you aren't "ranking" them, you are seeing if they allowed/scored more or less points; you are just trying to determine which team is better. If you have ever read my points for/against material, you will know that I don't really bother excluding defensive/STs scores because the team that scored deserves credit and the team that allowed it deserves to be downgraded.

 

[Watson's IQ]

Sorry but I cannot honestly reply. One hears scoring offense and defense bantered around alot in the press. I dont know if it is more valid than other metrics. My gut feeling is that it is more valid than yards gained and yards given up as it is the final score that gives the 'W'. But as you mentioned in another post, Football Outsiders uses some pretty advanced stats that probably could tell more. Unfortunately, those metrics are relatively new and therefore haven't been used for 37 seasons. So unless someone wants to go back all 37 seasons and apply the new statistics , we are left using the 'old traditional' ones. And that 0-37 sounds good to me!

You know what also sounds good to me: 19-0. Yeah Baby!

Haha, 19-0 sounds good to all of us. I think the problem with points scored and points allowed is that there's a lot of different ways a point can score, and fluky things happen over the course of a season. I agree that if you look at points allowed and points scored at the end of the season, you'd get a pretty good idea of who won the most games.

One stat I'm particularly fond of and have mentioned a couple times is the yards per point stat for defenses (it's more complex than I'm implying here, for simplicity), but basically how hard does your defense make the opponent work for each point (quantifies the bend but don't break thing, etc).

You're definitely right that football statistics are far from a perfect science yet (especially when compared to baseball), but if we're going to use them - and I think doing so would raise the level of discussion a lot - than we should try to use good ones and use them correctly.