Try Out the Updated Version of My QBTG Method

[Dan T.]

I hear Chucky and Lucky are both available.  Should we?  We could convert Chucky to tight end.

 

 

As to this latest OF rating system, as far as I can tell it is based entirely on random opinion and disregards any objective measure of individual performance.  But other than that, it's spot on.

[skinny21]

"Statistics involve objective data. I'm not sure what to call the product when you introduce subjective judgments." - OldFan

SKINNY: It seems to me that you could argue that rating a qbs mechanics etc uses a certain amount of subjectivity as well.

Absolutely. I made my point about your method only because you called the product "advanced statistics." Both methods, yours and mine, are subjective. On the other hand, I don't think an objective, statistical method is possible.

Do you feel/agree that it's possible to find similar situations for different qbs and rate their decision making in those instances? Perhaps with enough such instances we can come with a fair rating of football intellect.

I really don't know. You made me think of Patrick Ramsey at QB for Spurrier who wanted the ball put in the air deep to covered receivers even though his receivers didn't have the ball skills to outfight DBs as he wanted them to. If you were counting those throws against him, Ramsey would be unfairly punished. If you didn't count them at all, it would seem like Ramsey never made a mistake throwing deep.

Spurrier didn't protect his QBs well either. QBs under pressure are going to make more mistakes than well-protected QBs. Even if you forgive some of Ramsey's mistakes he's going to look worse than the well-protected QB because you are going to apply your standard evenly and forgive some of the well-protected QB's mistakes as well.

I haven't figured out how to separate quotes on my phone, so...

Regarding subjectivity and stats, look no further than tackles, hurries and sacks - a person tallying those often has to choose from more than one player that was in on the sack/tackle.

Regarding Ramsey, I'm talking about situations where all else is equal (as best as possible). I'd look at how many times the QB threw into double coverage when not under duress and rate him (and other qbs) accordingly. I would (for this particular stat) ignore plays that were outliers, such as hail mary's, end of game heroic type throws, possibly goal line throws, screen passes with multiple defenders in the area, etc.. The point is to find similar situations for different qbs to rate decision making.

Again, throwing into double coverage is just one example. How often do they hit a wide open receiver (assuming ample time to make their reads)? How often do they throw the ball away instead of stepping out for a loss? I'm no football guru, but I'm assuming a scout could come up with many more examples of situations where a QB can be separated from their supporting casts to rate their decision making.

Also, I would give the qbs the benefit of the doubt when I (as a scout) can't be sure of what I'm seeing.

[Oldfan]

Burgold: It's interesting enough, but the flaw in it for me is when is someone supposed to obtain the data to make the scoring.  If we can't use performance/games because of all of the variables introduced, when do we collect data?

 

 

You watch him play in NFL games and you grade as a scout might.

 

Footwork in a game is bound to be impacted by pressure.  If not in a one game isolation certainly over a season.  You can see the footwork of a player change over time.  Take David Carr or Pat Ramsey for example. 

 

 

Maybe you are more perceptive than I am. I didn't see the footwork of Carr or Ramsey change over time. I pointed out that Brady had a bad tendency to throw off his back foot when he felt pressure. That's a common fault. Lots of QBs do it. 

 

How do you grade accuracy?  A beautifully thrown spiral that misses its target could be due to a receiver running a route at incorrect depth, a bad throw, pressure, etc.

 

 

You watch the QB throw a bunch of passes when he isn't under pressure. You watch the flight of the ball and exactly how it arrives. Is it easy to catch? Does the receiver have to break stride? Was it thrown into a small window?  

 

Jason Campbell could toss the ball a mile, but rarely could toss it to where the receiver could catch it.  Is that a good arm or a bad arm?

 

 

Obviously accuracy is more important. But the ability to throw deep with touch is a valuable asset.

 

Aside from that, Liekert scales can be problematic.  Even with very, very precise operational definitions of what Excellent means or Poor means, you are going to get an incredible degree of variability.

 

 

The scale of five is just an easy way to communicate an opinion. It also allows one to use a simple formula in which weight has been assigned to other factors. Here, I assigned weights to the types of QBs.

[Mahons21]

 

Mahons: I believe is his ability to rapidly process information under duress. But I don't expect you to just take my word for it,

 

If Peyton's mechanics are strong why does he throw so many ducks?  I haven't seen Fitzpatrick as often as I've seen Peyton, but Peyton has a nice fade and I can't recall seeing Fitz ever throw one. I'd give Peyton an edge as a passer.

 

Favre had awful mechanics and an extremely tight spiral.  What do you not like about Peyton's mechanics?

 

 

 

Which explanation should one prefer? One that may or may not exist, or one which obviously exists and offers a cause that explains the effect? I opt for the latter. The big gap between Peyton and Ryan is that Peyton played in the same excellent scheme for 14 years and Ryan was never given that opportunity. If he had, the gap would have narrowed considerably.

 

Except, Peyton Manning has more or less been the same QB since his 2nd year in the system.  Which directly contradicts that his time in the system, is the advantage he's had v Fitz.

[Oldfan]

DanT: As to this latest OF rating system, as far as I can tell it is based entirely on random opinion and disregards any objective measure of individual performance.  But other than that, it's spot on.

 

Let's say that we want to compare Philip Rivers and Jay Cutler to determine which is the  better QB.  So, our question is: Which is the better QB?

 

The QBR formula is an objective measure. It would tell us that the San Diego Chargers passing game with Rivers at QB was better than the Chicago Bears passing game with Cutler at QB... but that doesn't answer our question.

 

Which objective measure of individual performance would you suggest? Which one am I disregarding?

[Jumbo]

So, to those whose main content is to simply note how the OP's "beginning proposal" includes much subjective (though, by the model hopefully well-informed) judgment (which he obviously noted up front), or is stillborn for other reasons, my question is are there any other models you like better, or how do you rate QBs? Mainly just with your eyes, taking in what others and media say, and "conventional" stats (as would seemingly be the case with the vast majority of fans)? 

 

 

Do you think there's even a point in trying to design a specific "system" by which you could dependably evaluate and rank a QB (or other positions, for that matter) in a standard, useful, application?

 

Is this model just a poor one in your opinions? I think Dan T. and Bang did a drop-in just to give that answer for themselves, so we know where they stand on OF (apparently ) , let alone this topic. How about some other efforts? Like our friends at Football Outsiders?

 

One value I find with a well-thought out effort at a model, is that if nothing else, it stimulates thinking on how to better evaluate such complex matters. While mainly just being a naysayer (especially when noting the "obvious" things) can have a valid (if very limited) use in this kind of exercise with this kind of beginning proposition, I am more entertained and given to thought by those who consider trying to add useful ideas, or adjust the ones presented by the OP to the way they think it would work better.

 

Personally, I am not really drawn to such formulaic approaches as a fan in this matter, but I do know that NFL FO's all over the league use various forms of such in varying manners and keep some of them quite protected. I know that some of the most sought after abilities among some supporting FO staff members are in research and statistical analysis, and development of models and formulas that show predictive validity in any number of matters, as opposed to "just going with your gut/instinct/knowledge" and "conventional" stats. That stuff always stays with us and still counts the most with people making the final call, of course. All this is (easily) arguably so, with both objective and subjective dynamics as part of an analysis.

[Dan T.]

DanT: As to this latest OF rating system, as far as I can tell it is based entirely on random opinion and disregards any objective measure of individual performance. But other than that, it's spot on.

Let's say that we want to compare Philip Rivers and Jay Cutler to determine which is the better QB. So, our question is: Which is the better QB?

The QBR formula is an objective measure. It would tell us that the San Diego Chargers passing game with Rivers at QB was better than the Chicago Bears passing game with Cutler at QB... but that doesn't answer our question.

Which objective measure of individual performance would you suggest? Which one am I disregarding?

I understand the problem you are trying to solve - that it's difficult to judge quarterback performance and compare them devoid of the talent and schemes surrounding them. But I don't see how you've solved that riddle with this method.

I also think you tend to overplay those factors (surrounding talent and scheme) when you downgrade quarterbacks that conventional wisdom says are good and elevate quarterbacks that conventional wisdom says are mediocre. You might be the only guy on the planet who dismiss Peyton Manning and Tom Brady as system quarterbacks. I won't be out on that limb with you.

Edit in light of Jumbo's post: Where I stand on this quarterback rating system is wildly different from where I stand on OF himself. I'm a huge OF fan. He gets real football discussions going around here. And he can take the heat from those who disagree, giving it right back.

So to summarize:

OF's QBTG = Cokkamamie

OF himself = The Man

[Oldfan]

MAHONS: Favre had awful mechanics and an extremely tight spiral.

 

Farve did not have awful mechanics. He used unconventional mechanics at times. Like Jay Cutler, he could do things when throwing a football that would be a mistake for lesser athletes to try.  

 

What do you not like about Peyton's mechanics?

 

You mean aside from the fact that his mechanics produce an ample supply of ducks? Well, there's not much to complain about other than that because there's not much footwork involved with throwing from the Shotgun. There are two basic ways to do it. Both are simple and easy to learn.

 

Except, Peyton Manning has more or less been the same QB since his 2nd year in the system.  Which directly contradicts that his time in the system, is the advantage he's had v Fitz.

 

Just a glance at his career numbers contradicts your point. I'm too lazy to do the math for you, but the average of his first five years will be significantly lower than from year six on. 

[Mahons21]

 

MAHONS: Favre had awful mechanics and an extremely tight spiral.

 

Farve did not have awful mechanics. He used unconventional mechanics at times. Like Jay Cutler, he could do things when throwing a football that would be a mistake for lesser athletes to try.  

 

Not but a few posts up you say throwing off your back foot is a bad tendency for QBs, now you're saying Favre (king of throwing off his back foot) was "unconventional.  While, something can be both a bad tendency and unconventional, you're choice of words seem highly subjective.

 

 

 

What do you not like about Peyton's mechanics?

 

You mean aside from the fact that his mechanics produce an ample supply of ducks? Well, there's not much to complain about other than that because there's not much footwork involved with throwing from the Shotgun. There are two basic ways to do it. Both are simple and easy to learn.

 

If you don't think Peyton has some of the best footwork in the NFL, we agree to disagree.  You don't need to perform a drop back to have strong footwork, it's also how you move your feet within the pocket when going through reads, and how you manipulate the pocket to deal with pressure, and Peyton as at the tops of the league in this regard.

 

Re: the Ducks

If a tight spiral was the intent of Peyton's throwing mechanics, I have full confidence he could achieve it.  However that is not his intent.  Peyton's intent is to deliver an accurate and catchable pass.  I imagine if he didn't use touch passes as much as he does, you would see much tighter spirals, as they are an indication of a ball with more velocity.

 

 

Except, Peyton Manning has more or less been the same QB since his 2nd year in the system.  Which directly contradicts that his time in the system, is the advantage he's had v Fitz.

 

Just a glance at his career numbers contradicts your point. I'm too lazy to do the math for you, but the average of his first five years will be significantly lower than from year six on. 

 

Isn't your whole premise of this thread that stats are misleading?  I'm telling you that since the mistakes he made his rookie season, Peyton Manning has more or less been the same QB throughout his career.  Increased yards/TDs can be attributed to a shift in the scheme away from rushing and towards more passing, different supporting cast (the emergence of Reggie Wayne as a bonafide #1).

[Oldfan]

Jumbo: How about some other efforts? Like our friends at Football Outsiders?

 

LKB suggested I check out Football Outsiders long ago. I've now been a member there for years. I was especially impressed with their candor on individual statistics. They pointed out, for example, what Martin and I have said in this thread about QB statistics. They tell us what how good the San Diego passing game is with Philip Rivers at QB. They don't tell us how good Philip Rivers is.

 

Sadly, that candor is gone from the site. I'm disappointed with them.

[Dan T.]

The best field research into this topic is when a quarterback switches teams or when a new quarterback comes into an existing team.  That gives you a pretty good comparative snapshot.  Peyton Manning from the Colts to the Broncos has passed this test. RG3, handed the keys to a lemon of a Redskins offense, passed this test.  Jay Cutler passed the test.

 

Jason Campbell, bounced from the Redskins to the Raiders to...where the hell is he now? has not passed this test.  Neither did Donovan McNabb.

[Oldfan]

Dan: I understand the problem you are trying to solve - that it's difficult to judge quarterback performance and compare them devoid of the talent and schemes surrounding them. But I don't see how you've solved that riddle with this method.

 

I would much prefer an objective method, but there isn't one. I'm going to assume you agree since you did not answer my question: 

Which objective measure of individual performance would you suggest? Which one am I disregarding?

 

You might be the only guy on the planet who dismiss Peyton Manning and Tom Brady as system quarterbacks. I won't be out on that limb with you

.

 

The view from the limb is often very good.

 

Sam Huff taught me that all football players are system players. Sam looked like an absolute terror playing in the Giants scheme surrounded by their talent. He came to the Skins in a trade and looked ordinary playing for us.

 

All players are system players and that includes QBs. Knowing that enabled me to predict that Donovan McNabb would fail here. Donovan was misfit for any scheme other than the one Andy Reid built around him.

 

Mark Schlereth offered the only intelligent explanation for why Shanahan traded for him. He thinks that Mike believed he could correct the mechanical flaw in McNabb's game that made him inconsistent on short-to-intermediate throws. That's possible. 

 

OF's QBTG = Cokkamamie

 

 

Boooo!

 

OF himself = The Man

 

 

Yeaaaa!

[Dan T.]

Huff had several really good years for the Redskins, even though he was 30 and on the downside of his career by the time he came to DC. 

 

The raft of standard stats compiled for quarterbacks, coupled with observation, go further than any attempt to quantifying out surrounding talent and scheme. 

 

and I can't wrap my head around your assertion that all players are system players,

[Oldfan]

Mahons: Not but a few posts up you say throwing off your back foot is a bad tendency for QBs, now you're saying Favre (king of throwing off his back foot) was "unconventional.  While, something can be both a bad tendency and unconventional, you're choice of words seem highly subjective.

 

Cutler also throws off the back foot. It isn't a problem for Cutler and Farve because they can do it and still get off an accurate pass while doing it. In fact, it's an advantage because, since they don't need to set their feet, they can get the pass away quicker.

 

The reason they can do it and conventional QBs can't is because, in my opinion, conventional mechanics, those generally taught to young QBs, are flawed approaches.

 

I've read that RG3's father had him throwing from a sitting position while Robert was recovering from his first knee surgery. That takes the legs out of it and produces a more compact throwing motion that will allow Robert to throw well even without a solid base. He could throw off his back foot like Favre or Cutler when a quicker throw is needed.

 

If you don't think Peyton has some of the best footwork in the NFL, we agree to disagree.

 

Never a problem.  I enjoy our exchanges more when we disagree.

 

If a tight spiral was the intent of Peyton's throwing mechanics, I have full confidence he could achieve it

.  

Not a chance.

 

Isn't your whole premise of this thread that stats are misleading?

 

 

ALL stats? No. Most football stats are misused.

 

The QBR is a pretty good quick-and-dirty stat for grading a team's passing game or its defense's effectiveness in stopping the opponent's passing game. The Skins D had an 80 rating--ranked 18th in 2012, BTW.

 

The QBR is also useful in grading Peyton's effectiveness in his scheme from year-to-year.

[Oldfan]

Skinny, when OCs grade their QBs, they first debrief them. What did they see from their field level position? What didn't they see? What coverage did they read? The third-read receiver got open; did he have time to get to his third read? Then, armed with the knowledge of what the QB was supposed to do, the OC can give a fair grade.

 

Anyone trying to grade the QB as you have suggested would have to do it without that debriefing and without the expertise the OC has in his own scheme.

 

Your method makes sense as far as you have explained it. I think you might be able to produce something useful, but it would be a huge, time-consuming process.

[Oldfan]

Dan: The best field research into this topic is when a quarterback switches teams or when a new quarterback comes into an existing team.  That gives you a pretty good comparative snapshot.  Peyton Manning from the Colts to the Broncos has passed this test.

 

Peyton moved to Denver and brought his scheme with him. Denver provided better players than Peyton left behind in Indy. That's not much of a test. .

 

Jeff Garcia started for four different teams. His career QBR stats show the volatility that  different schemes and supporting casts produce.

 

In one season, Tom Brady had the worst WR corps in the NFL. In the next, he had the best. His QBR increased by 33% just on improvement of just that one element.

[Oldfan]

Dan: and I can't wrap my head around your assertion that all players are system players,

 

 

The word "synergy" was coined to describe the effect when two metals are fused and the product has better qualities than either of the metals used. Good cooperation (teamwork) has a synergistic effect. The performances of everyone: players, coaches, front office personnel will never be as good elsewhere. That's why both Belichik and Brady are overrated and it's why the reputations of Pioli and others who left the Pats organization lost their lustre.

 

The reverse is true also. The personnel leaving poorly functioning organizations might well be underrated. You have to be able to see such talent apart from their losing background -- like a London Fletcher. Fletcher didn't receive the recognition he deserved because he spent most of his prime years with Buffalo.

[skinny21]

Skinny, when OCs grade their QBs, they first debrief them. What did they see from their field level position? What didn't they see? What coverage did they read? The third-read receiver got open; did he have time to get to his third read? Then, armed with the knowledge of what the QB was supposed to do, the OC can give a fair grade.

Anyone trying to grade the QB as you have suggested would have to do it without that debriefing and without the expertise the OC has in his own scheme.

Your method makes sense as far as you have explained it. I think you might be able to produce something useful, but it would be a huge, time-consuming process.

True, it would be time consuming (not something you and I have the time for anyway). I would tend to think a scout tasked with watching every snap a QB takes could make a note of what they see in terms of football decisions (when applicable) and then come up with the percentages to grade different qbs (on their football decisions).

This strikes me as the sort of thing an outside source (Football Outsiders) might be inclined to do. It sure would be interesting to see if these things (along with a rating system like you've come up with) jive with the 'eye test'.

[Dan T.]

 

Dan: and I can't wrap my head around your assertion that all players are system players,

 

 

The word "synergy" was coined to describe the effect when two metals are fused and the product has better qualities than either of the metals used. Good cooperation (teamwork) has a synergistic effect. The performances of everyone: players, coaches, front office personnel will never be as good elsewhere. That's why both Belichik and Brady are overrated and it's why the reputations of Pioli and others who left the Pats organization lost their lustre.

 

The reverse is true also. The personnel leaving poorly functioning organizations might well be underrated. You have to be able to see such talent apart from their losing background -- like a London Fletcher. Fletcher didn't receive the recognition he deserved because he spent most of his prime years with Buffalo.

 

I think we have a different notion of what a "system" player means.

[Oldfan]

Dan: I think we have a different notion of what a "system" player means.

 

Probably. That label was originally pinned on college QBs who played in pass-dominant schemes which inflated their stats. I expanded it because it was a deceptive label. All player performances are both scheme-dependent and dependent on the quality of the players in their unit. That's why, no matter how flawed, the "scouting method" I'm advocating here is the only sensible way to grade talent at any position. 

[Burgold]

I think if I were to come up with a scheme to rate QB's they'd have to include these factors

 

Read:  Does the qb make the correct throw and understand what the defense is doing.  How many dumb/dangerous throws do they make?  (The weight of this is based on your own preference for risk/reward)

 

Accuracy:  Does the ball arrive at a place where the receiver can catch it?  This might also include is it thrown in a place where the defender's opportunity to deflect or intercept the ball is limited.

 

Touch:  Does the QB throw the ball with the appropriate velocity to hurt the defense and aid the receiver?

 

Mechanics:  Is his throwing technique, footwork, etc. appropriate?

 

Armstrength: Can he execute the throws the plays demand?  How far can he hurl the ball?

 

Mobility: This includes both escabability, that knack for sliding around in the pocket as well as the qbs agility and foot speed.

 

I think most of these are quantifiable.  Which categories are most important probably depend more on scheme.  Walsh might rate armstrength as a middle priority and be much more concerned with touch and accuracy.  The great Raiders teams might believe that mechanics and armstrength are the end all/be all.  Several might think that read is very high up there (IE knowing what the defense is about, really understanding what and why the play has been called and what is happening and likely to happen)

[Oldfan]

Burgold: I think if I were to come up with a scheme to rate QB's they'd have to include these factors

 

 

I agree with everything you said except the first factor.

 

Read:  Does the qb make the correct throw and understand what the defense is doing.  How many dumb/dangerous throws do they make?  (The weight of this is based on your own preference for risk/reward).

 

 

If we're hoping to come up with something that a well-informed NFL fan could use, we can't include "read." This is an intangible area that would require far more information and expertise than a fan could possibly have.

 

If you go back to the opening of my OP, and the identical twins, Chucky and Lucky. Most fans would credit Lucky with a fantastic ability to read and react to defenses and downgrade his twin even though they are identical in this respect. What the media and fans do is to use the intangibles to explain the difference in performances which are actually explained by the scheme and talent supporting the QB. It's just one effect of overrating the QB position.

 

Skinny offered up a theory that, with an enormous amount of effort and expertise, some of those intangibles might be quantified. If a tool like that became available, I'd use it.

[Burgold]

Possibly, but we might be able to see/intuit if the qb checks correctly.  How many times does a qb call an audible that results in a positive play for instance.  We, as fans, certainly can tell the difference between a qb who is a risk taker and a game manager.  Say a Brett Farve who would very often try to squeeze a ball in there versus a Gibbs era Brunnell who almost never threw a ball to a receiver unless he was wide open and was content to throw it away/check down on any down and distance.

[Oldfan]

Burgold: Possibly, but we might be able to see/intuit if the qb checks correctly.  How many times does a qb call an audible that results in a positive play for instance.  We, as fans, certainly can tell the difference between a qb who is a risk taker and a game manager.  Say a Brett Farve who would very often try to squeeze a ball in there versus a Gibbs era Brunnell who almost never threw a ball to a receiver unless he was wide open and was content to throw it away/check down on any down and distance.

 

 

Yes, certainly, over the course of time we can see/intuit patterns such as you suggest. But who are you trying to grade, the QB or the coach? In most cases you will be grading the coach's scheme.

 

Some coaches won't allow the QB to call an audible. They play the percentages. They have a large playbook; they scout their opponent's defense to determine what defenses they are likely to use in a given situation; and they pick plays that are likely to succeed. Other coaches create a small playbook and rely on having more practice time on each play in order to maximize precision. Then they allow the QB to audible at the LOS depending on how they read the safeties.

 

Some coaches, like Joe Gibbs, are paranoid about turnovers. They choose game managers like Brunell. Others, can live with talented gunslingers. If this forum is a reliable gauge, I'd guess that most fans would grade in a way that favored the game manager because he would throw fewer interceptions.

 

How many fans do you suppose understand that you can't discuss INTs unless you discuss TDs in the next breath? The gunslinger will put up more of both. TDs are usually worth seven points. The average INT is a negative four. So, if the gunslinger puts up an extra TD for every INT, he's +3 on the board for each.

 

The QB's support system, coaching, offense, defense and special teams, will also have an effect on his risk-taking. If he's playing with a poor support system, he will be playing the come-from-behind game more often than his counterpart with a good support system. That means more risk taking and , consequently, more "mistakes."

[skinny21]

Yes, certainly, over the course of time we can see/intuit patterns such as you suggest. But who are you trying to grade, the QB or the coach? In most cases you will be grading the coach's scheme.

Some coaches won't allow the QB to call an audible. They play the percentages. They have a large playbook; they scout their opponent's defense to determine what defenses they are likely to use in a given situation; and they pick plays that are likely to succeed. Other coaches create a small playbook and rely on having more practice time on each play in order to maximize precision. Then they allow the QB to audible at the LOS depending on how they read the safeties.

Some coaches, like Joe Gibbs, are paranoid about turnovers. They choose game managers like Brunell. Others, can live with talented gunslingers. If this forum is a reliable gauge, I'd guess that most fans would grade in a way that favored the game manager because he would throw fewer interceptions.

How many fans do you suppose understand that you can't discuss INTs unless you discuss TDs in the next breath? The gunslinger will put up more of both. TDs are usually worth seven points. The average INT is a negative four. So, if the gunslinger puts up an extra TD for every INT, he's +3 on the board for each.

The QB's support system, coaching, offense, defense and special teams, will also have an effect on his risk-taking. If he's playing with a poor support system, he will be playing the come-from-behind game more often than his counterpart with a good support system. That means more risk taking and , consequently, more "mistakes."

Yeah, the 'eye test' tells us a lot, particularly those that have a great deal of experience, but as you point out, there's much more involved than just the qbs play.

I try to set aside the bigger plays (positive or negative) when evaluating a player, but that's far from foolproof. It's one of the reason I never bought into the talk about players like Manning (pre super bowl win) and McNabb not being able to 'win the big games'. Besides the fact that they'd both won multiple do-or-die playoff games, there's no telling if the defenses stepped up, the supporting casts made more mental mistakes, or the coaches made their own mistakes in a pressure-packed situation.

It's also why, when looking at stat lines for runningbacks for fantasy football, I throw out the longest and shortest runs to get a slightly more accurate snapshot of their game.