Fantasy Football - Footballguys Forums
This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!
Greatest post-season performances by a QB (1 Viewer)
- Thread starter Chase Stuart
- Start date
David Yudkin
Footballguy
Are you looking at just the totals for each particular post season and factoring the number of games each played? Delhomme played 4 games when most other guys only played 3. Who can forgot Delhomme's huge game of 9 completions for 101 yards against the Eagles that year. I realize his 3 other games were pretty stout, but having an extea game didn't hurt his numbers any.
Chase Stuart
Footballguy
It's not totals; it's adjusted yards above the league average. So if Delhomme threw an extra 10 passes for an extra 50 yards, he'd actually be hurt by those numbers.And while you might rip on Delhomme's game against the Eagles, it's not like he had many attempts. That game was pretty insignificant for him in constructing his rating.David Yudkin said:
dirty_martini
Footballguy
Good work - thought provoking.
Just curious - Can you expound upon this statement from your site? "To be frank, I also feel that for the most part, post-season performances are overrated. Guys get way too much credit for post-season wins and losses, in my opinion, and it really overshadows the talent or ability of the QB." Isn't that what this game is all about? Winning the championship?
Just curious - Can you expound upon this statement from your site? "To be frank, I also feel that for the most part, post-season performances are overrated. Guys get way too much credit for post-season wins and losses, in my opinion, and it really overshadows the talent or ability of the QB." Isn't that what this game is all about? Winning the championship?
radballs
Footballguy
Well, there is 21 other guys on the field so a QB can be really good without "winning the championship", i.e. Dan Marino. The opposite is also true, i.e. Trent Dilfer.Also, there is still the issue with substantially small sample sizes when it comes to evaluating most, if not all, QBs based on their historical post season performances.
GregR_2
Footballguy
Actually, I think your statement is a great example of exactly what Chase is talking about. He's discussing the quality, talent and ability of Quarterbacks. But then you say, "but the game is about winning the championship". If Chase was discussing the most successful teams in NFL history, you would be correct that should play a big role in the discussion.But he's not, he's talking about how good the quarterbacks are. Winning a championship and the quality of your QB are not completely unrelated... but neither are they so closely tied that you can correctly infer something about a QB just by the number of championships he has. Football is enough of a team game that it's very possible the best QB in the league might never make the championship game let alone win one.As another way of illustrating this, most QBs who won a championship had at least one game in his postseason that was decided by a small enough margin, a single play he wasn't even involved in could have swung the game the other way and cost his team their championship. For example, a RB who fumbles it and the other team runs it back in could have cost this guy the game.The point is, when rating a QB, how heavily should you rate a championship when you know that he might not have won it had other players on his team done differently... and thus knowing that other QBs out there may have won that championship instead if their teamate had performed as this guy's did? Surely you can rate the playoff game as at least equal to a regular season, and maybe a little more important. But I think Chase is arguing people give championships too much weight when judging a QB, compared to the role the QB plays in attaining them.
Last edited by a moderator:
dirty_martini
Footballguy
Well the name of the thread and his article is Greatest post-season performances by a QB. Losing the game makes it very tough for your performance to be called the greatest - and that applies to all players on the team not just the QB. You make solid points about other members on the team could have done things differently that influenced the outcome of the game. Not going to argue with that at all.
But the QB is one that is called upon to be a leader. The team winning or not reflects greatly on the ability of the leader on the team to lead. Maybe the safety blowing a coverage to give up a late TD is the most important play of the game up to that point. What the leader of the team coaxes out of his teammates from that point forward might make it just another play in the game ...
On the other side of the coin - A QB's stats might be hurt by sound football decisions. Say throwing the ball away instead of forcing a risky pass or taking a sack when a RB heads the wrong way. Or switching out of a passing play because the defense is perfect for a running play.
I don't know how to do it - but I think winning should be reflected in the rankings.
Chase Stuart
Footballguy
Forcing a risky pass that leads to an INT is going to be penalized severely in my system. Unfortunately, I don't have post-season sack data yet, but I think I made it clear in the post that I think that's a big problem. I do consider sack data when looking at the regular season numbers, because I already have that data.The question of whether "winning the game" is important is distinct from who was the best QB. They're related, of course, but different.
Sidewinder16
Footballguy
Interesting that it took him more than twice as many games to edge out Plunkett in order to get in the top 5.These stats are blatantly false. Everyone knows that John Elway is a big fat choker. Any list that includes him in the top 5 is a total sham.
BigSteelThrill
Footballguy
Thanks again Chase. You get a big fat 5-star rating for these recent statistical offerings. 
Tau837
Footballguy
First off, I didn't have a chance to read the full blog post yet. Just chose a few examples to look at quickly. First, I am assuming that the year referenced is the year of the regular season (e.g., Simms' season cited above is for the postseason following the 1986 regular season). I guess this is obvious, but it can get slightly confusing when the playoff games have a different year cited.So the three years I chose above contain these aggregate performances:
Favre 1995:
WildCard; Sun Dec 31, 1995; W, GNB 37, ATL 20; 24/35 (68.6%), 199 passing yards (5.7 ypa), 3 TDs, 0 interceptions, 111.5 rating
Division; Sat Jan 6, 1996; W, GNB 27, SFO 17; 21/28 (75%), 299 passing yards (10.7 ypa), 2 TDs, 0 interceptions, 132.9 rating
ConfChamp; Sun Jan 14, 1996; L, GNB 27, DAL 38; 21/39 (53.8%), 307 passing yards (7.9 ypa), 3 TDs, 2 interceptions, 84.0 rating
Total: 66/102 (64.7%), 805 passing yards (7.9 ypa), 8 TDs, 2 interceptions
Favre 1996:
Division; Sat Jan 4, 1997; W, GNB 35, SFO 14; 11/15 (73.3%), 79 passing yards (5.3 ypa), 1 TD, 0 interceptions, 107.4 rating
ConfChamp; Sun Jan 12, 1997; W, GNB 30, CAR 13; 19/29 (65.5%), 292 passing yards (10.1 ypa), 2 TDs, 1 interception, 107.3 rating
SuperBowl; Sun Jan 26, 1997; W, GNB 35, NWE 21; 14/27 (51.9%), 246 passing yards (9.1 ypa), 2 TDs, 0 interceptions, 107.9 rating
Total: 44/71 (62%), 617 passing yards (8.7 ypa), 5 TDs, 1 interception
To me, given that you aren't accounting for winning in your ranking system, Favre's 1995 performance is intuitively better than his 1996 performance, yet you rank the 1996 performance higher.
Simms 1986:
Division; Sun Jan 4, 1987; W, NYG 49, SFO 3; 9/19 (47.4%), 136 passing yards (7.2 ypa), 4 TDs, 0 interceptions, 111.0 rating
ConfChamp; Sun Jan 11, 1987; W, NYG 17, WAS 0; 7/14 (50%), 90 passing yards (6.4 ypa), 1 TD, 0 interceptions, 94.3 rating
SuperBowl; Sun Jan 25, 1987; W, NYG 39, DEN 20; 22/25 (88%), 268 passing yards, 3 TDs, 0 interceptions, 150.9 rating
Total: 38/58 (65.5%), 494 passing yards (8.5 ypa), 8 TDs, 0 interceptions
Simms grades out with a very efficient performance, but it is carried by one game. (Well, I know the 4 TDs in the division game also mattered.) His Super Bowl performance is possibly the best postseason QB performance ever, but the other 2 games are underwhelming, especially when you consider the 4 TDs in the division game came in a 49-3 blowout. While not as efficient, I'd prefer Favre's overall performance in 1995 to Simms' group of 3 games in 1986. And I think I'd say the same for 1996, though it is a tougher call.
ConstruxBoy
Kate's Daddy
Maybe I'm just a Bills homer, but it's hard to take this seriously when it asserts that Jim Kelly is the 3rd worst playoff QB of all time. It just doesn't match what I and many others saw in those games. Especially when it looks to be based on one Super Bowl in which he's killed by the number of passes he attempted. I think you need to adjust that someway.
Chase Stuart
Footballguy
Well, I'm open to suggestions. But:1) I do think Super Bowl performances should be weighted. People don't talk about Peyton Manning's game against the Broncos the way they do Montana's game against the Broncos. There's a reason for that.2) "Worst" ever lists when comparing to the league average are tricky. But Kelly stinking for a lot of attempts shouldn't be better than Kelly stinking for a small number of attempts.Playoff data is based on really small sample sizes, so I'm not of the opinion that Jim Kelly couldn't have had some terrific playoff games if he had more chances. But when he did play, he was pretty far below average in the most important games. He does rank 3rd worst, but keep in mind exactly what that's saying. That's all.
PlasmaDogPlasma
Footballguy
How do you discredit a starting QB for TDs he threw in a blow-out? Part of the reason the game was a blow-out was Simms throwing those TDs.Anyway, I'll take the zero post-season interceptions for Simms over either year for Favre.
Tau837
Footballguy
I didn't discredit Simms. But the game was over long before he threw the last 2 TD passes against the 49ers (it was 28-3 and Montana had been knocked out of the game when Simms threw his third).Part of Chase's methodology here is to give a 10 yard bonus for TDs and a 45 yard penalty for interceptions. Favre's 805/8/2 is equivalent to 795 yards; his 617/5/1 is equivalent to 622 yards; Simms' 494/8/0 is equivalent to 574 yards.
I was hoping Chase would comment on this to explain what it is that elevates Simms' performance above the two Favre performances, especially since this posted list weighs all playoff games equally (i.e., the Super Bowl is not weighted more than the other games) and winning/losing is not a factor.
Chase Stuart
Footballguy
The fact that Simms got his yards on much fewer attempts.
Tau837
Footballguy
It's not quite that simple. Favre in 1996 averaged 8.7 ypa, compared to Simms' 8.5. I assume this has to do with the conversion from YPA to AYA, but thought perhaps you could illuminate. This one just doesn't pass the eyeball test to me.
The other thing is that this makes me question a bit what the right balance is between quantity and quality, where quality is measured by efficiency.ETA: And the other problem with this particular ranking is that all of the individual ranks are based on such a small number of games, it is easy to take context into account (e.g., my comments about Simms throwing TDs in a blowout), which doesn't happen with your system. Not that your system could or should account for this, I just would expect there to be more issues with this particular list than the lists based on full seasons or careers.
Last edited by a moderator:
Chase Stuart
Footballguy
You're correct that it's in the conversion. I'll go through it in a second. But your bigger point is spot on, and why it's pretty tough to get a meaningful ranking of post-season performances. For the most part, the data isn't big enough to produce any reliable results.That said, in 1986, the league average QB averaged 5.59 AY/A; in the '86 playoffs, Simms averaged 9.90 AY/A. Over 58 attempts, we'd expect an average QB to put up 324 adjusted yards; as you said, Simms put up 574 adjusted yards. Thus, Simms averaged 250 adjusted yards above average. He also had some noteworthy rushing accomplishments -- in the first game he had 1 carry for 15 yards. Not impressive, sure, but that's 11 rushing yards over 4.0 YPC. In the SB, he had 3 carries for 25 yards...that's another 14 rushing yards over 4.0. So combined, Simms had 250 passing yards over average and 24 rushing yards over average, making him +274 for the playoffs.It's not quite that simple. Favre in 1996 averaged 8.7 ypa, compared to Simms' 8.5. I assume this has to do with the conversion from YPA to AYA, but thought perhaps you could illuminate. This one just doesn't pass the eyeball test to me.The other thing is that this makes me question a bit what the right balance is between quantity and quality, where quality is measured by efficiency.
ETA: And the other problem with this particular ranking is that all of the individual ranks are based on such a small number of games, it is easy to take context into account (e.g., my comments about Simms throwing TDs in a blowout), which doesn't happen with your system. Not that your system could or should account for this, I just would expect there to be more issues with this particular list than the lists based on full seasons or careers.
Favre, in the '96 playoffs, had 622 adjusted yards on 71 pass attempts. The league average that year was 5.54 AY/A, so we'd expect 394 adjusted yards on 71 attempts. Thus, Favre was +228. Favre, in the SB, had 22 adjusted yards on 4 rushing attempts, giving him 6 AY over average. 228 + 6 = 234. So that's how Simms, on 58 pass attempts, comes up to have better stats than Favre, on 71 attempts.
The shorthand is that Favre averaged 8.8 AY/A on slightly more attempts, while Simms averaged 9.9 AY/A. Simms' advantage is quality slightly outweighed Favre's edge in quantity. But this stuff is just an approximation -- it's tough to draw any conclusions based on five or six games, much less two or three.
ConstruxBoy
Kate's Daddy
I don't really feel like arguing this, to be honest. But statistics that don't match perception a little bit can be useful and insightful. Statistics that don't match perception by a large amount are sort of useless to me. I just think, looking at the results, that this one falls in the latter category. Neil O'Donnell lost that Super Bowl for Pittsburgh with two horrible interceptions to Larry Brown, among other poor decisions. I would bet that unbiased viewers who watched that game and the Kelly Super Bowl thought that O'Donnell had the worst game, by quite a bit. But your statistic has Kelly 250 "points" worse than O'Donnell? I just don't think that matches reality, which admittedly is much harder in football than in baseball because there are so many external factors involved.
Chase Stuart
Footballguy
To be clear, Kelly was 250 yards worse than O'Donnell for the entire post-season, not for the Super Bowl.Kelly does rank lower than O'Donnell in the Super Bowl, but I agree that O'Donnell is given more blame for the loss. Probably because his team lost a close game where he threw 3 INTs, and Kelly lost a blowout where he threw 4 INTs and a ton of incompletions.Anyway, I wouldn't put too much weight on statistical analysis of QB's based on one playoff game. The simple version is generally, throwing 58 attempts and 275 yards and 2 TDs/4 INTs is less impressive than 49 attempts, 239 yards and 1 TD/3 INTs. That's the "more often than not, performance two is better than performance once." But over the sample size of one game, that often doesn't hold up. (And in this example, O'Donnell only gets a slight win over Kelly; his better performance in the earlier rounds is why he's ahead of him by a few hundred yards.)
ConstruxBoy
Kate's Daddy
So you're saying the traditional way of looking at QB stats (in which I think almost everyone would put Kelly ahead of O'Donnell there because of more yards, more TDs and only 1 more pick) is wrong and that attempts are more important? Or is it completion attempts? Y/A? I just don't see it. And thanks for clearing up the playoff season thing. I missed that.
Chase Stuart
Footballguy
I think for one game, stats in one game are pretty unreliable.But you can't just say Kelly had more yards without recognizing that he had more attempts. And one more interception is a bit worse than one more touchdown. 300 yards, 3TD and 3 INT isn't as good as 300 yards, 0 TD and 0 INT, on the same number of attempts. Kelly and O'Donnell's performances were pretty similar. They were both bad. And in general, being bad for 60 attempts is worse than being bad for 50 attempts, just like being bad for 50 attempts is worse than being bad for 5 attempts. I think the methodology works much better when finding the best QBs by season, than the worst QBs by game. But I still think even in the latter case, it's right more often than not (as opposed to right very often).
Brownsfan07
Footballguy
why don't they add kelly holcomb's 449 yards on the squeelers so-called vaunted defense?
if andre king hadn't of been tackled in bounds, holcomb would've had another 300+ yard performance guaranteed!
just my 2 cents!
if andre king hadn't of been tackled in bounds, holcomb would've had another 300+ yard performance guaranteed!
just my 2 cents!
Tau837
Footballguy
This is not true in all cases. Using your system of 10 yards per TD and -45 yards per interception, a TD that is 35 yards or longer has more positive value than the negative value of an interception.
I disagree that you can draw this conclusion based on these facts alone. First of all, suppose it took the 300/0/0 QB 60 attempts, and it took the 300/3/3 QB only 15 attempts. Pretty unrealistic, but we just don't know based on the limited facts presented.Aside from that issue, what do we know about these two performances given only the stated facts? We know the 300/3/3 QB produced 3 TDs for his team, and at worst all 3 of his interceptions were returned for TDs... But that is unlikely, and it is very possible that not all of the interceptions led to TDs for the opponent, in which case the 300/3/3 QB directly produced a positive point differential for his team. We don't know anything about the scoring in the 300/0/0 performance, either for the QB's team or its opponent, so we cannot know his performance produced a positive point differential for his team.
Last edited by a moderator:
bostonfred
Footballguy
Not sure I believe that Brady's 2003 performance, in which he outdueled both of the co-MVPs in terrible weather, then won a shootout in the Superbowl against the #2 QB on your list when the Patriots secondary was decimated by injury, is worse than both Manning and Delhomme's performances during that same year.
Chase Stuart
Footballguy
As I said in the post, playoff data is very murky to judge based on statistics because of small sample sizes. Things like weather don't even out over the course of two games, so Brady doesn't get a "boost" in this system for out-dueling MVPs by throwing an incredible two touchdowns in two games. Manning had two absolutely incredible playoff games that year, and Delhomme had an oustanding post-season. Both of them had some historic games in those post-seasons.
Chase Stuart
Footballguy
One more interception, coming with the same number of passing yards for the game, is worse than one more TD. So a 40 yard TD adds a touchdown and 40 yards to a QB's totals.60 and 15 aren't the same number of attempts.
Tau837
Footballguy
For some reason I missed the same number of attempts qualifier even though I quoted it.
That eliminates my middle point. But I think my first and last point stand.
Chase Stuart
Footballguy
Either AY/A or net adjusted yards per attempt, compared to the league average, are really good tools (IMO) for grading QB seasons across eras. Otto Graham's 1953 comes out pretty close to Tom Brady's 2007. Bert Jones' 1976 falls between Sid Luckman's 1943 and Steve Young's 1992. That's pretty cool, and I also think that's a fair and accurate representation of the quality of the seasons.Using it for a single game, of course, is very risky. I don't have a ton of confidence behind it, but: a) it's better than nothing; and b) it's better than anything else I can think of to compare every QB playoff game over the course of 40 seasons.For some reason I missed the same number of attempts qualifier even though I quoted it.That eliminates my middle point. But I think my first and last point stand.
I'm always interested in hearing alternative systems. I made a good chunk of changes between the 2006 and 2008 editions of my QB study.
bostonfred
Footballguy
While the Colts certainly needed all 304 yards and all three of his TDs in the 38-31 victory over the Chiefs, giving him credit for a "historic game" there seems like a bit of a reach. It was a good game in the divisional round, but it pales in comparison to Brady's shootout in the Superbowl the same year. And while he blew up against Denver, I'm sure Colts fans would agree that they'd rather he'd saved a couple of the 5 TDs he threw in a 41-10 rout of the Broncos for the game against New England, where he threw four picks. Statistics will say that he threw for an incredible 10 TDs and only four INTs in those playoffs while averaging over 300 yards per game, but the reality is that he threw a bunch of garbage TDs against a Denver team that didn't show up on either side of the ball, beat a defenseless Chiefs team, and then got crushed. I can buy the argument that you're averaging out their performances, but you also have Manning's 2004-5 postseason on there, in which Manning again blew up against Denver (this time for four TDs) but was held to no TDs and an interception against the Pats the following weekend. Why is that postseason performance was better than Brady's 2004 postseason, where he had 2 TDs per game and threw no picks? They both averaged 2 TDs per game, but Manning threw for more yards but also threw a pick. It's also interesting to note that Brady has better postseason numbers overall than Manning. Manning has more yards on a per game basis, but Brady has more TDs, fewer INTs, and a better passing percentage, and has been much more consistent. 16 of Manning's 24 career TDs came in four of his 14 postseason games. He's been held without a TD in three. Brady has never been held without a touchdown in a complete playoff game (he was knocked out in the first half of the Pittsburgh game but was 12 of 18 for 115 yards when he left). In short, your system rewards guys who have a big game and a terrible game over guys who play consistently better. The results seem to bear out that consistency is better, which is why Delhomme and Manning appear on the list while several Superbowl winners do not.Chase Stuart said:
Similar threads
