PDA

View Full Version : The Stats Don't Lie



NorthCoast
11-19-2009, 11:30 PM
These are Ben's career splits for wins-losses. It looks to me like when we ask Ben to do too much it is not a good outcome. It leads to INTS and low yards per attempt. We need a balanced O.


http://i904.photobucket.com/albums/ac248/bubbarmw/statscopy.jpg

stlrz d
11-19-2009, 11:54 PM
Stats don't ever tell the whole story. Those numbers are completely out of the context of the games in which they occurred.

You can't just look at the volume stats of a QB and make a declaration like, "When he's asked to do too much it is not a good outcome."

Jason Whitlock summed it up best when he said baseball is a game of statistics and football is a game of relying on your eyeballs.

SteelerNation1
11-20-2009, 12:40 AM
I'd venture to say that manning, Brady, Montana, whoever had worse #'s in games they lost rather than games they won.

RuthlessBurgher
11-20-2009, 01:23 AM
I'd venture to say that manning, Brady, Montana, whoever had worse #'s in games they lost rather than games they won.

And you'd be correct! How about that! (sorry my formatting is not as good as the O.P.)

Tom Brady

Value G W L T Cmp Att Cmp% Yds TD Int Rate Y/A AY/A A/G Y/G

Win 93 93 0 0 1976 3036 65.1% 23459 184 50 101.9 7.73 8.20 32.6 252.2

Loss 29 0 29 0 558 969 57.6% 5726 32 42 67.6 5.91 4.62 33.4 197.4


Joe Montana

Value G W L T Cmp Att Cmp% Yds TD Int Rate Y/A AY/A A/G Y/G

Win 124 124 0 0 2381 3690 64.5% 29833 228 80 101.1 8.08 8.34 29.8 240.6

Loss 53 0 53 0 1028 1701 60.4% 10718 45 59 73.1 6.30 5.27 32.1 202.2

SteelAbility
11-20-2009, 08:50 AM
Stats don't ever tell the whole story. Those numbers are completely out of the context of the games in which they occurred.

You can't just look at the volume stats of a QB and make a declaration like, "When he's asked to do too much it is not a good outcome."

Jason Whitlock summed it up best when he said baseball is a game of statistics and football is a game of relying on your eyeballs.

Over large number of games the individual game-influencing factors and context "come out in the wash." Some for you, some against you. The larger the sample set, the more that stuff balances out. Statistics can lie in the short run. But in the long run they are a very good indicator. Not the whole story. But a very good indicator.

SteelAbility
11-20-2009, 08:55 AM
I'd venture to say that manning, Brady, Montana, whoever had worse #'s in games they lost rather than games they won.

I believe NorthCoast's point is that Ben is passing significantly more in the losses ...

24.9 Attempts per game in the wins
31.9 Attempts per game in the losses

Now you could argue that he's passing more because either the run support isn't there. or the D gave up a lot of point so we had to find ways to lengthen the game ... by passing more. Sure, I'll buy into that.

JTP53609
11-20-2009, 09:23 AM
i totally agree, based on stats, when we ask ben to do alot, often times it results in a loss....

well exept last year in the super bowl when he marched us 85 yards down the field for the championship,
or that game in baltimore when he drove us 92 yards,
or that home game against dallas when he led us down to tie the game with under 3 minutes left
or the san diego game and jacksonville games from last year when we relied on his arm to win the game....i wont even mention 2007 when he threw for 32 touchdowns and had all those comeback from behind wins that season, or in the playoffs in 2005 against cincy, indy and denver when he was nearly perfect...but yea other than that, when we ask him to do alot he usually comes up empty for us and we are screwed, i say lets draft someone

stlrz d
11-20-2009, 09:33 AM
I'd venture to say that manning, Brady, Montana, whoever had worse #'s in games they lost rather than games they won.

I believe NorthCoast's point is that Ben is passing significantly more in the losses ...

24.9 Attempts per game in the wins
31.9 Attempts per game in the losses

Now you could argue that he's passing more because either the run support isn't there. or the D gave up a lot of point so we had to find ways to lengthen the game ... by passing more. Sure, I'll buy into that.

That's no argument...it's the truth! That's why the stats, even over the "long run", do lie. They don't account for any of those things...they are just numbers.

frankthetank1
11-20-2009, 09:40 AM
well any qb is going to throw more in games there team is behind in.

Oviedo
11-20-2009, 09:45 AM
I'd venture to say that manning, Brady, Montana, whoever had worse #'s in games they lost rather than games they won.

I believe NorthCoast's point is that Ben is passing significantly more in the losses ...

24.9 Attempts per game in the wins
31.9 Attempts per game in the losses

Now you could argue that he's passing more because either the run support isn't there. or the D gave up a lot of point so we had to find ways to lengthen the game ... by passing more. Sure, I'll buy into that.

Could it be we are behind in losses and therefore have to throw more to catch up. Isn't that what every QB would have to do if behind in a game? As correctly stated--totally out of context.

Accept the fact that this is a league that has legislated significant advanategs for the passing game, most of our offensive talent are those players who directly contribute to the passing game, we just paid Ben $102M to do more than hand off the ball. We will pass more from here on out regardless of what stats say.

papillon
11-20-2009, 09:47 AM
Stats don't ever tell the whole story. Those numbers are completely out of the context of the games in which they occurred.

You can't just look at the volume stats of a QB and make a declaration like, "When he's asked to do too much it is not a good outcome."

Jason Whitlock summed it up best when he said baseball is a game of statistics and football is a game of relying on your eyeballs.

MY eyeballs tell me that having a ratio of 70% pass plays to running plays is a losing proposition for the Steelers almost every time. They can probably get away with that against KC, Cleveland and a few others, but even average teams will beat them if the Steelers don't have better balance.

Do you believe that allowing Ben to air it out 40 times and rush the ball 18 times is a good mix of play calling for the Steelers to be successful? Those stats are a bit telling and it wasn't until last year and this year that Ben was actually able carry the team with his arm on occasion. If the Steelers begin to rely on Ben too heavily they will fail.

Pappy

JUST-PLAIN-NASTY
11-20-2009, 10:04 AM
These are Ben's career splits for wins-losses. It looks to me like when we ask Ben to do too much it is not a good outcome. It leads to INTS and low yards per attempt. We need a balanced O.


http://i904.photobucket.com/albums/ac248/bubbarmw/statscopy.jpg

Wow...I fell like this is something MrSmartMonies (I think that was his name.) Would post. I'm not trying to push you out of the way or anything...But looking at those stats says that when Ben has a bad game the Steelers don't win. Not the Steelers lose when Ben has to pass alot.

Further Stats:

Att per game
25 Att-W
32 Att-L

TD per Att
17 Att/TD-W
27-Att/TD-L

INT per Att
41 Att/Int-W
19 Att/Int-L

To Make you argument, you would need to bring up Bens stats when he throws say more than 25 times in a game. He has done that every game this season so...

9-G 205/Com 302/Att 67.9/% 2,469/Yds 8.2/Avg 14/TD 8/Int 97.1/Rate

Splits By Wins/Losses G Att Comp Pct Yds Avg Lng TD Int 1st 1st% 20+ Sck SckY Rate
Losses 3 106 65 61.3 671 6.3 51 2 3 33 31.1 8 7 51 74.1
Wins 6 196 140 71.4 1,798 9.2 52 12 5 87 44.4 24 20 144 109.6


Att per game
33 Att-W
35 Att-L

TD per Att
16 Att/TD-W
53 Att/TD-L

INT per Att
39 Att/Int-W
35 Att/Int-L

I don't see how you come to that conclusion. Ben is throwing the ball almost the same times a game W or L. Career versus this year, you can make an argument it was because his attempts were lower when he first came into the league. His attempts this year exceed his career attempts when he has a losing record. I think the biggest thing you could take from these stats is the consistancy of attempts to INts regardless of outcome. He actually throws less Ints per attempt when they lose. The biggest thing that sticks out is his TDs per attempt in W & L. To me, that says the Steelers lose if Ben can't get them in the endzone. He doesn't even average a TD a game in their Ls with 2 more attempts. Ben averages 2 Tds a game when they win. The saying still stands with the Steelers...They always have a Top 5 defense year to year...But the Steelers only go as far as Ben can take them. When Ben plays well...The Steeler win. Not when Ben throws alot the Steelers lose.

stlrz d
11-20-2009, 10:06 AM
Stats don't ever tell the whole story. Those numbers are completely out of the context of the games in which they occurred.

You can't just look at the volume stats of a QB and make a declaration like, "When he's asked to do too much it is not a good outcome."

Jason Whitlock summed it up best when he said baseball is a game of statistics and football is a game of relying on your eyeballs.

MY eyeballs tell me that having a ratio of 70% pass plays to running plays is a losing proposition for the Steelers almost every time. They can probably get away with that against KC, Cleveland and a few others, but even average teams will beat them if the Steelers don't have better balance.

Do you believe that allowing Ben to air it out 40 times and rush the ball 18 times is a good mix of play calling for the Steelers to be successful? Those stats are a bit telling and it wasn't until last year and this year that Ben was actually able carry the team with his arm on occasion. If the Steelers begin to rely on Ben too heavily they will fail.

Pappy

Where did I state that the pass/run ratio against the Bengals was ok?

The point is the numbers over Ben's entire career taken out of context are not a fair way to simply say, "When he's asked to do too much we are in trouble" because for many of those games (I know you remember them as well as I do) we were in a hole and we needed the guy to throw the ball a ton of times...which for any QB is a challenge. Look at what RB posted and you'll see.

Northern_Blitz
11-20-2009, 10:11 AM
When you're winning, you are more likely to run to kill the clock. When you're not, you are more likely to pass to try to put up points quickly. It's not rocket science.

Look at defensive passing yards per game. Washington is the best in the league, and Carolina is in the top 5. Do you think that's because their passing Ds are awesome, or because they aren't that good and teams don't try to pass on them in the second half of games?

papillon
11-20-2009, 10:13 AM
[quote="stlrz d":3celwyxf]Stats don't ever tell the whole story. Those numbers are completely out of the context of the games in which they occurred.

You can't just look at the volume stats of a QB and make a declaration like, "When he's asked to do too much it is not a good outcome."

Jason Whitlock summed it up best when he said baseball is a game of statistics and football is a game of relying on your eyeballs.

MY eyeballs tell me that having a ratio of 70% pass plays to running plays is a losing proposition for the Steelers almost every time. They can probably get away with that against KC, Cleveland and a few others, but even average teams will beat them if the Steelers don't have better balance.

Do you believe that allowing Ben to air it out 40 times and rush the ball 18 times is a good mix of play calling for the Steelers to be successful? Those stats are a bit telling and it wasn't until last year and this year that Ben was actually able carry the team with his arm on occasion. If the Steelers begin to rely on Ben too heavily they will fail.

Pappy

Where did I state that the pass/run ratio against the Bengals was ok?

You didn't and I know that, but an implication of that was in your response (at least in my feeble mind), so I ran with it. :Cheers

The point is the numbers over Ben's entire career taken out of context are not a fair way to simply say, "When he's asked to do too much we are in trouble" because for many of those games (I know you remember them as well as I do) we were in a hole and we needed the guy to throw the ball a ton of times...which for any QB is a challenge. Look at what RB posted and you'll see.

And many times those holes were dug by Ben himself in the early years. It does become a challenge for any quarterback to have to throw it often without the threat of a run to retake control of a game or even win it in that situation. Regardless, of why Ben has to throw it an abnormal amount of time (whether by his own hand or some other facet of the game) the Steelers are by and large unsuccessful when he does.

[/quote:3celwyxf]

Pappy

JUST-PLAIN-NASTY
11-20-2009, 10:24 AM
When you're winning, you are more likely to run to kill the clock. When you're not, you are more likely to pass to try to put up points quickly. It's not rocket science.

Look at defensive passing yards per game. Washington is the best in the league, and Carolina is in the top 5. Do you think that's because their passing Ds are awesome, or because they aren't that good and teams don't try to pass on them in the second half of games?

If you are refering to the Bengals game...The Steelers had the lead...They were tied...And never were further back than one posession. Those stats should have been as close to balance as you could get. Something to the effect of 53% pass to 47% run at most because the Steelers had 2 posessions at the end trailing.

If you take away Ben's scramble...The Steelers set up to pass 71% of the plays and ran the ball 29% of the time. You are right...It isn't rocket science.

feltdizz
11-20-2009, 11:43 AM
40 passes and 17 runs in a game we never trailed by more than one score is a bad ratio.

SteelAbility
11-20-2009, 12:14 PM
I'd venture to say that manning, Brady, Montana, whoever had worse #'s in games they lost rather than games they won.

I believe NorthCoast's point is that Ben is passing significantly more in the losses ...

24.9 Attempts per game in the wins
31.9 Attempts per game in the losses

Now you could argue that he's passing more because either the run support isn't there. or the D gave up a lot of point so we had to find ways to lengthen the game ... by passing more. Sure, I'll buy into that.

That's no argument...it's the truth! That's why the stats, even over the "long run", do lie. They don't account for any of those things...they are just numbers.

The wording I put out was poor. Stats CAN, in fact, lie in the long-run. It boils down to this. The longer the run the lower the likelihood of a lie. The possibility of uncertainty is always there, unless you are talking about the laws of physics or some other unchanging principle.

For example, if you took what you were told was a "fair coin" and flipped it 10 times (short run) and it came up heads twice, you wouldn't question strongly that it was fair. If you flipped it 10,000 times and it came up heads 2,000 times you would question very strongly that it is fair. Nonetheless IT IS POSSIBLE, in the 10,000 flip scenario that it really is a fair coin. What's the likelihood though? Same thing with player stats. The longer the run, the higher the likelihood of concluding the "real truth." Sometimes a RB gets good blocking, sometimes not. Over several thousand carries, all of that comes out in the wash and his YPC is converging on a number that makes a strong indication about his true running ability.

By the way, I did a quick calculation of the odds of a fair coin coming up with 20 heads in 100 tries (scaled down from 10,000). Statisticians say that at about 100 sample points, statistics start to become very significant. It came out to about 4.2 in 10 Billion.
So, in the 20/100 scenario there is a 4.2 in 10 Billion chance that the "fair coin" claim is true.

SteelTorch
11-20-2009, 01:10 PM
These are Ben's career splits for wins-losses. It looks to me like when we ask Ben to do too much it is not a good outcome. It leads to INTS and low yards per attempt. We need a balanced O.


http://i904.photobucket.com/albums/ac248/bubbarmw/statscopy.jpg
EDIT: Nevermind.

Secondly, did it ever occur to anyone that MAYBE Ben has to throw more because he's throwing badly to begin with, not the other way around?

NWNewell
11-20-2009, 01:50 PM
I agree that stats don't lie... but they sure as hell can be mis-represented.. as these numbers are.

This is like the argument that when teams run more, they win. When the pass more, they lose. Why are teams running more? Often it's because they are already head and are trying to grind out the game. But running wasn't "the reason" that they were ahead and won. Why teams do teams that pass more often lose? Because they are behind and need to make up ground quickly with chunks of yards. So they try to pass more. but passing is not necessarily what made them fall behind and lose.

Same can be said for Ben's numbers. When we are ahead we may run more to grind out the clock and he has the luxuray of using the whole play bock and making high percentage plays. But when we are behind, we need to make up ground quickly and he has to pass more... and at that, he has to take more risk to move down the field quickly. So obviously, with a smaller list of plays to choose form and having to take more risk, his succuess rate is going to be less.

We won't even get into other variables like no run support for a passing game no passing support for the running game, or vise versa. Or one dimensional predictable play calling, or other elements of the teams play as well as their opponents.

As an Engineer I have a lot of faith in statistical analysis. But what some of you guys call statistics is far, far from it. You might as well pull a number out of a random number generator and say "See, lookie here! This confirms my predisposition!!"

As stlrs d said, those numbers are completely out of context and mis-represented.

PS I'm not trying to argue that you are wrong. Just that your evidence that you are right is bogus. We definitely need a balanced offense. No offense is going to have much success when they have 70/30 splits. I don't care if that 70 percent is running OR throwing. But to say that we ask Ben to throw the ball 33 times, we are more likely to loose than if we throw the ball 27 times is silly.

SteelAbility
11-20-2009, 02:13 PM
I agree that stats don't lie... but they sure as hell can be mis-represented.. as these numbers are.

This is like the argument that when teams run more, they win. When the pass more, they lose. Why are teams running more? Often it's because they are already head and are trying to grind out the game. But running wasn't "the reason" that they were ahead and won. Why teams do teams that pass more often lose? Because they are behind and need to make up ground quickly with chunks of yards. So they try to pass more. but passing is not necessarily what made them fall behind and lose.

Same can be said for Ben's numbers. When we are ahead we may run more to grind out the clock and he has the luxuray of using the whole play bock and making high percentage plays. But when we are behind, we need to make up ground quickly and he has to pass more... and at that, he has to take more risk to move down the field quickly. So obviously, with a smaller list of plays to choose form and having to take more risk, his succuess rate is going to be less.

We won't even get into other variables like no run support for a passing game no passing support for the running game, or vise versa. Or one dimensional predictable play calling, or other elements of the teams play as well as their opponents.

As an Engineer I have a lot of faith in statistical analysis. But what some of you guys call statistics is far, far from it. You might as well pull a number out of a random number generator and say "See, lookie here! This confirms my predisposition!!"

As stlrs d said, those numbers are completely out of context and mis-represented.

PS I'm not trying to argue that you are wrong. Just that your evidence that you are right is bogus.

You put your finger on a key question, i.e. cause/effect. This is where stats don't tell the whole story. However, I'm not ready to necessarily pin it on "passing because we are losing" or "losing because we are passing." It's a two-edged sword and it swings both ways. The truth is somewhere in the middle.

For example, we all complained in the 07 season when we played Denver and we came out all cylinders passing when Denver was known to be one of the league's weakest run defenses. IMO, that was definitely a case of "losing because we were passing." If we had stuck to the run game to begin with, we wouldn't have got caught behind and had to do all that passing (which incidentally involved a Ben sack and fumble returned for a TD).

eniparadoxgma
11-20-2009, 04:06 PM
well any qb is going to throw more in games there team is behind in.

$$$$$$

The OP's stats suggest a correlation, which is not the same thing as a causal relationship. This is Stats 101 stuff.

What team doesn't throw more when they are behind?

What if our run game was shut down in the majority of those games?

What happens when the defense knows you're going to throw?

I don't understand the point of taking a statistic out of context and trying to interpret it to mean a causal relationship when only a correlation can be found.

On edit, I didn't see NWNewell's until after I posted. Good stuff.

Again, you cannot just take a specific stat like this and assume that there is a causal relationship. There are too many variables at play, (a lot of which have already been mentioned).

feltdizz
11-21-2009, 01:06 AM
So stats lie now and fans know the truth? lol....

I understand each game plays out differently but I think this post is alluding to the obvious... We abandon the run way too early... We just watched the game so the numbers aren't lying to us... 40 passes when you are only down one possession the whole game is pathetic.