Cowboys Stupid Mistake Rating, Week 5

[fanfromvirginia]

At the beginning of the season, I came up with an idea for a new stat, which I'm calling the Stupid Mistake rating. The formula is giveaways x 6 + # of penalties.

This week the Cowboys had 2 giveaways and 13 penalties for a rating of 25. This brings their 5-game average to 17.0,the highest it's been this season. The 25 was the 2nd worst performance this year, with the NYG rating at 28 (4 giveaways and 4 penalties). We have been dropping the last couple weeks and now are squarely in between last year's league median and our last-in-the-league rating of 20. Last year we averaged 2.1 giveaways/game (worst in the league) and about 7.5 penalties/game (also worst in the league). This year we've been a bit and up and down so far but, again, the trend is in the wrong direction.

League best 2008 (the Patriots): 9.0
League median 2008: 15.1
Cowboys 2008 (league worst): 20.0
Cowboys 2009: 17.0

You can see the link below for this rating but, in short, it's not designed to explain why we're losing; its only purpose is to be an easy way to keep track of and compare our stupid mistakes over time and against other teams. It is NOT ideal -- it is designed to be easy to compute and compare so it leaves out all sorts of helpful detail that would be very hard to keep track of league-wide (e.g., calling a timeout to avoid a delay of game penalty or, say, calling two fade routes in a row on 2nd and goal from the one yard line on a night you're averaging 8 yards per rush ).



(For details see http://cowboyszone.com/forums/showthread.php?t=161334)

 

[poke]

i thought after yesterdays game you might have to build a new meter to measure the stupid mistakes.

 

[casmith07]

Still doesn't account for what constitutes a giveaway vs a takeaway for the other team. If the other team makes a miraculous play for an INT, how does that constitute a stupid mistake? Also, I feel like penalties should be weighted depending on yardage.

I'm on my phone right now but I'll post a formula that I think would provide a more accurate rating scale. You need more variables.

 

[ZeroClub]

I'm not sure there is a number that can capture the stupidity that was on display in Kansas City.

 

[The Quest for Six]

i thought after yesterdays game you might have to build a new meter to measure the stupid mistakes.

well, atleast one for Flozell Adams

 

[AbeBeta]

Still doesn't account for what constitutes a giveaway vs a takeaway for the other team. If the other team makes a miraculous play for an INT, how does that constitute a stupid mistake? Also, I feel like penalties should be weighted depending on yardage.

I'm on my phone right now but I'll post a formula that I think would provide a more accurate rating scale. You need more variables.

Why is the present approach it necessarily inaccurate? Unless you expect any one team to be victimized by great defensive plays more than another, your argument is not valid.

 

[Future]

stupid mistakes are easy to rate...who gives up the most yards without a play being run. that's dallas.

What did they have 8 penalties before the snap? Ridiculous.

 

[fanfromvirginia]

Still doesn't account for what constitutes a giveaway vs a takeaway for the other team. If the other team makes a miraculous play for an INT, how does that constitute a stupid mistake? Also, I feel like penalties should be weighted depending on yardage.

I'm on my phone right now but I'll post a formula that I think would provide a more accurate rating scale. You need more variables.

Sounds good. Keep in mind that I want something that's easy to compute. What's interesting about number of penalties v. penalty yardage is that, on the surface, there doesn't seem to be too much difference. Dallas was last year in one category (if memory serves) and second to last in the other. So I chose to use penalties because it's easier to compute.

That's the key for me -- easy to compute that combines turnovers and penalties (both of which we've been bad at the last couple years and both of which constitute, in general, stupid mistakes).

I totally agree that it is not ideal.

 

[fanfromvirginia]

Why is the present approach it necessarily inaccurate? Unless you expect any one team to be victimized by great defensive plays more than another, your argument is not valid.

Yeah, my two key considerations with this were to make it easy to compute and compare, which it unquestionably is, and to make it a rough, general measure of stupidity.

When you start getting into subjective categories you've got to spend a lot of time watching film or find someone else who has already done something like this.

 

[AbeBeta]

When you start getting into subjective categories you've got to spend a lot of time watching film or find someone else who has already done something like this.

even worse... with a stat like this you will end up just about where you started.

 

[Chocolate Lab]

League best 2008 (the Patriots): 9.0
League median 2008: 15.1
Cowboys 2008 (league worst): 20.0
Cowboys 2009: 17.0

Do you know what the league average is for this year?

 

[fanfromvirginia]

Do you know what the league average is for this year?

Coming soon. I'll work it out sometime before, during or after the bye-week and have it ready for the next update. League wide, penalties seem to fluctuate from year to year at least somewhat so it's a good idea I think.

 

[DallasDomination]

There was a point in that games yesterday where I thought surely the Cowboys were playing a practical joke on the sports world. No way you make this many stupid mistakes and have no clue where to line up. I dont know how many times I saw brookin tell his guys where to line up.

The Coaching on this team is either that stupid and bad or we have some equally ******** players.


Thank god Romo played Good and Brookins kept the D in Shape...And thank god we played the Chiefs.

 

[Q_the_man]

At the beginning of the season, I came up with an idea for a new stat, which I'm calling the Stupid Mistake rating. The formula is giveaways x 6 + # of penalties.

This week the Cowboys had 2 giveaways and 13 penalties for a rating of 25. This brings their 5-game average to 17.0,the highest it's been this season. The 25 was the 2nd worst performance this year, with the NYG rating at 28 (4 giveaways and 4 penalties). We have been dropping the last couple weeks and now are squarely in between last year's league median and our last-in-the-league rating of 20. Last year we averaged 2.1 giveaways/game (worst in the league) and about 7.5 penalties/game (also worst in the league). This year we've been a bit and up and down so far but, again, the trend is in the wrong direction.

League best 2008 (the Patriots): 9.0
League median 2008: 15.1
Cowboys 2008 (league worst): 20.0
Cowboys 2009: 17.0

You can see the link below for this rating but, in short, it's not designed to explain why we're losing; its only purpose is to be an easy way to keep track of and compare our stupid mistakes over time and against other teams. It is NOT ideal -- it is designed to be easy to compute and compare so it leaves out all sorts of helpful detail that would be very hard to keep track of league-wide (e.g., calling a timeout to avoid a delay of game penalty or, say, calling two fade routes in a row on 2nd and goal from the one yard line on a night you're averaging 8 yards per rush ).



(For details see http://cowboyszone.com/forums/showthread.php?t=161334)

Tell me one thing is this your idea or someone else. you have the total from 2008 Patriots and league avg. since you have the league avg from last year, can you please post all teams to see where we compare or at least our division, thanks.

 

[fanfromvirginia]

Tell me one thing is this your idea or someone else. you have the total from 2008 Patriots and league avg. since you have the league avg from last year, can you please post all teams to see where we compare or at least our division, thanks.

It was my idea but I never computed the team by team averages. I took the easy way out for 2008. The Patriots were first in both categories, so they had to be first overall. We were last in both categories so we had to be last. I didn't take the league average -- I took the league median, which in this case is halfway between the 16th and 17th team. That was also easy to find for both stats and didn't require compiling and computing all the data.

So, in short, I don't have those figures. What I'm going to do over the next couple weeks is to compile all the info so far for each team for this season in an Excel spreadsheet so it can be easily compiled from here out. At that point I'll make the type of info you're looking for available.

On edit: We were actually 2nd to last in turnovers last year and last in penalties but last overall in my rating.

 

[casmith07]

Since the boneheadedness of a penalty rises with the yardage usually, I say you value them each as such to total a 1.0 variable:

5-yarders: 0.15
10-yarders: 0.25
15-yarders: 0.60

Multiply each variable by the number of penalties. For a boneheaded giveaway, i.e. a grossly bad interception or a fumble, count it as a 1. For a good defensive play, a la the first Giants pick or the scoop-off-the-turf by Champ Bailey last week, multiply those by .25. Divide the whole thing by the total amount of plays from scrimmage.

[(FiveYarders*.15)+(TenYarders*.25)+(FifteenYarders*.60)] + [(Giveaways)+(Takeaways*.25)] / Plays From Scrimmage

I think - while the variables and constants are completely erroneous - this formula will give you something that at the very least weights the different situations fairly across the league in order to get a more accurate number. With your current rating formula, a team that gets 3 false starts is just as liable as a team that gets 3 unsportsmanlike conduct penalties...not a fair assessment. The giveaways vs. takeaways is purely subjective as well, but at least the formula won't penalize a team for the other team making a great play.

Giveaway = throwing a pick into triple coverage
Takeaway = throwing a pick into one-on-one man coverage that's tipped by the DB and snagged by the safety

I'll do up an Excel spreadsheet to help you out :

 

[fanfromvirginia]

Since the boneheadedness of a penalty rises with the yardage usually, I say you value them each as such to total a 1.0 variable:

5-yarders: 0.15
10-yarders: 0.25
15-yarders: 0.60

Multiply each variable by the number of penalties. For a boneheaded giveaway, i.e. a grossly bad interception or a fumble, count it as a 1. For a good defensive play, a la the first Giants pick or the scoop-off-the-turf by Champ Bailey last week, multiply those by .25. Divide the whole thing by the total amount of plays from scrimmage.

[(FiveYarders*.15)+(TenYarders*.25)+(FifteenYarders*.60)] + [(Giveaways)+(Takeaways*.25)] / Plays From Scrimmage

I think - while the variables and constants are completely erroneous - this formula will give you something that at the very least weights the different situations fairly across the league in order to get a more accurate number. With your current rating formula, a team that gets 3 false starts is just as liable as a team that gets 3 unsportsmanlike conduct penalties...not a fair assessment. The giveaways vs. takeaways is purely subjective as well, but at least the formula won't penalize a team for the other team making a great play.

Giveaway = throwing a pick into triple coverage
Takeaway = throwing a pick into one-on-one man coverage that's tipped by the DB and snagged by the safety

I'll do up an Excel spreadsheet to help you out :

I appreciate the effort and agree that you would need your system or something like it to get a truly accurate measure of stupidity. But I'm sticking with my method because it's easy. All you need is a box score. With your method you'd have to watch each and every game in the NFL pretty carefully in order to compare against other teams. And there would be some subjectivity.

I'm not sure I'd want to weight the penalties that way, anyway. 15 yarders can be pretty stupid and they're certainly more harmful but some of these 5 yarders are really needless as well. A lot of ten yard penalties are holding calls, many of which are OL making mistakes in tough situations.

I think that in general giveaways and penalties involve some stupidity and they're really, really easy to keep track of. I'm making the assumption (admittedly a rather large one) that the various discrepancies even out enough that this is still a worthwhile way to compare the Cowboys across seasons and against other teams. Is it ideal? No, but neither is batting average in baseball and that's been around for more than a century.

 

[goshan]

Nice job on this.
Your focus on something easy to compute is dead on.
It is always tempting to add more complexity, but it would likely provide marginal value and it would take a ton of time for someone, meaning it won't get done.

 

[casmith07]

well after all of that, the spreadsheet is done. All it needs is numbers you're talking to an engineer here -- efficiency is key!

I might change it up to make it easier to locate data, but I'm sure there's a place out there where these stats could be easily drawn from in order to compute the finals.

 

[casmith07]

Update: for all the non-math whizzes, I've simplified the spreadsheet to use box score stats. I'm in the process of updating it right now and will get the highs and lows for everyone...

For those wondering, the Chiefs out-dumbed us 9.65 to 14.63 using the a new and improved formula...details coming.