The CardBoard
Nice stat of the week - Printable Version

+- The CardBoard (https://thecardboard.org/board)
+-- Forum: C-House! (https://thecardboard.org/board/forum-4.html)
+--- Forum: The CardBoard (https://thecardboard.org/board/forum-5.html)
+--- Thread: Nice stat of the week (/thread-10044.html)


Nice stat of the week - Mick - 09-30-2015

Stanford has just three turnovers in the first four games.

Another good stat - Hulk01 - 09-30-2015

Our offense was mistake-plagued for most of last season.  Here are the comparative "bad play" stats for our games against USC and Oregon State over the last two years:

2014:
10 offensive penalties
3 sacks
8 fumbles (yes, eight)
2 passes intercepted
23 bad plays total
2015
6 offensive penalties
3 sacks
1 fumble
I pass intercepted
11 bad plays total

Re: Nice stat of the week - CTcard - 09-30-2015

(09-30-2015, 12:01 PM)Mick link Wrote:Stanford has just three turnovers in the first four games.

It appears to me that Stanford has GAINED just three turnovers in the first four games.
We seemed to have given up four (2 fumbles, both by McCaffrey and two interceptions) to put us a -1 in turnover differential for the year, #10 in the conference.

Re: Nice stat of the week - 82lsju - 09-30-2015

(09-30-2015, 12:21 PM)CTcard link Wrote:[quote author=Mick link=topic=12989.msg131794#msg131794 date=1443639717]
Stanford has just three turnovers in the first four games.

It appears to me that Stanford has GAINED just three turnovers in the first four games.
We seemed to have given up four (2 fumbles, both by McCaffrey and two interceptions) to put us a -1 in turnover differential for the year, #10 in the conference.
[/quote]

correct

http://www.gostanford.com/ViewContent.dbml?&DB_OEM_ID=30600&CONTENT_ID=1402063

Re: Nice stat of the week - OutsiderFan - 09-30-2015

Stanford used to be a turnover forcing and sack machine.  What the hell happened?

Re: Nice stat of the week - fullmetal - 09-30-2015

DL depth.  A fierce and effective pass rush can usually cause a number of turnovers.  With a thin DL, we can't collapse the pocket as quickly.  Also, teams may be running more on us (less prone to turnovers).  As the Stanford offense starts clicking, teams will be forced to pass more (to catch up), and our opportunities for turnovers will increase.  I expect Duane Akina's magic will cause our secondary to develop sticky hands as well--we've dropped a few surefire interceptions.

NB: David Parry had his first NFL sack this weekend.  Stanford produces legitimate players.

Re: Nice stat of the week - stupac2 - 09-30-2015

I was thinking about this, mostly in context of some analysis of some NFL team's ability to generate takeaways (I think it was the Jets, who are something stupid like +9 so far on the season). Generating takeaway opportunities is a skill (though the takeaways themselves are somewhat random, especially fumble recovery), and it seems like generating those opportunities could come at a cost. For instance, against OSU I kept wondering why we weren't jumping those slants that seemed to be the only thing they could throw, and I can only assume that there reason is we were betting that playing sound man-on-man coverage and using the extra defenders to commit to stuffing the run was a better option. If we went aggressively for interceptions and screwed up it could easily lead to a fast TD.

Someone who's better at understanding these tradeoffs might have more, but that's my take on it. It's possible that as trust in the secondary improves we'll move away from that. It's also possible that we'll start generating more pressure and creating better opportunities for INTs.

Re: Another good stat - Spiny_Norman - 09-30-2015

(09-30-2015, 12:13 PM)Hulk01 link Wrote:Our offense was mistake-plagued for most of last season.  Here are the comparative "bad play" stats for our games against U$¢ and Oregon State over the last two years:

2014:
10 offensive penalties
3 sacks
8 fumbles (yes, eight)
2 passes intercepted
23 bad plays total
2015
6 offensive penalties
3 sacks
1 fumble
I pass intercepted
11 bad plays total

To me, the starkest compare-and-contrast is how these "bad" plays show up on the scoreboard:

2014 vs USC
2 for 5 in the Red Zone

2015 vs. USC
6 for 6 in the Red Zone

Re: Nice stat of the week - OutsiderFan - 09-30-2015

Since interceptions are generally tied to QB pressure and Stanford's pass rush isn't generating too many sacks, that explains lack of interceptions a bit.  Off the top of my head I can think of four very close to sacks by the defense this year, so it's not like QBs are being given all day to throw either.

To me, forced fumbles are the measure of a team that is good at forcing turnovers.  If the defense forces fumbles, its generally aggressive.  Defenders are making concerted efforts to strip the ball, get their helmets on them, and when a ball carrier is wrapped up by a teammate, they attack they go for the ball.

What I notice from Stanford is an effort to play sound, fundamental defense.  Everyone does his job and nobody tries to be a hero leaving his responsibility to make a play he doesn't trust a teammate to make.  Stanford's defense keeps things in front and forces an offense to execute all the way down the field. 

It would be nice to see maybe a smidgen more risk taking, either send more blitzers, or a little more effort to attack the ball.  But hard to complain with the results the defense has produced to date.  There is no reason to take more risks on defense when the offense is scoring like it is.  Playing sound is enough to win games.

Re: Nice stat of the week - AlbanyCardinal - 09-30-2015

(09-30-2015, 12:46 PM)OutsiderFan link Wrote:Stanford used to be a turnover forcing and sack machine.  What the hell happened?

This doesn't tell the whole story, by any means, but how many times have we seen a CB or safety drop a sure int.? Seems to happen every game and going back a few years.

Re: Nice stat of the week - BostonCard - 09-30-2015

Turnovers are pretty random events, especially "forcing" turnovers.

http://www.footballoutsiders.com/stat-analysis/2004/guest-column-turnovers-and-unpredictability-defense

There is much less correlation from year to year on turnovers than most other stats.

BC

Re: Nice stat of the week - stupac2 - 09-30-2015

(09-30-2015, 02:50 PM)Boston Card link Wrote:Turnovers are pretty random events, especially "forcing" turnovers.

http://www.footballoutsiders.com/stat-analysis/2004/guest-column-turnovers-and-unpredictability-defense

There is much less correlation from year to year on turnovers than most other stats.

BC

But at the same time, at least for INTs, the rule is something like 2/9 passes defensed turn into INTs, and clearly getting passes defensed is a skill (as an aside, I really hate the phrase "passes defensed" but it seems to be the term of the art here so whaddayagonnado?). As the author of that article says, there are so many year-to-year changes in the NFL that it's not too surprising that the correlation is low. And you expect correlation on offense to be higher since QBs tend to move around less and avoiding INTs is definitely a skill. Even with coaching continuity, player changes could easily change these numbers, look at NE losing 3/4 of their top secondary players last year, naturally the scheme is going to change and that could affect their TO numbers. The sort of simple analysis in that post is not really going to be able to figure this out.

There's obviously still a lot of luck involved, but I think the general idea that teams can weight the dice in their favor (or against) with talent and scheme makes way too much sense to be dismissed so easily.

Re: Nice stat of the week - BostonCard - 09-30-2015

I think the author accounted for changes in personnel by looking at the year-to-year correlation in other statistics, finding that those are much higher than "forcing turnovers".  If every season were truly independent due to roster turnover, all the stats would be uncorrelated, but some, such as yards allowed, are more correlated than turnovers.

I think there is probably a better statistical way of answering the question, which is to look at turnovers within a season.  I assume that turnovers per game can be modeled with a poisson distribution (https://en.wikipedia.org/wiki/Poisson_distribution), and you can look at between team departures from that distribution that would exceed the within team departures.

By the way, we've had 1, 1, 0, and 1 turnovers in our four games.  If you model turnovers with a Poisson distribution, the 95% confidence interval of our turnover rate is 0.19 to 1.94 turnovers per game (that is, the observed data would be consistent with a long term rate of between 1 turnover every 5 games, in which case we have been really lucky, and 2 turnovers per game, in which case we have been unlucky).

BC

Re: Nice stat of the week - CowboyIndian - 09-30-2015

(09-30-2015, 03:42 PM)Boston Card link Wrote:  I assume that turnovers per game can be modeled with a poisson distribution...

BC

You mean like Jesus did?

Re: Nice stat of the week - stupac2 - 09-30-2015

(09-30-2015, 03:42 PM)Boston Card link Wrote:I think the author accounted for changes in personnel by looking at the year-to-year correlation in other statistics, finding that those are much higher than "forcing turnovers".

Well, sure, no one is arguing that turnovers aren't noisy. (Although, to be honest, I have no idea exactly what you're trying to argue here.) The question is "are turnovers completely random or is there skill in generating them?" To use your terms, can teams affect the lambda in the distribution? I think that the answer is obviously "yes", and it's just to what extent and what things they can do to affect it.

Quote:If every season were truly independent due to roster turnover, all the stats would be uncorrelated, but some, such as yards allowed, are more correlated than turnovers.

Again, who said everything is entirely independent? That's a terrible assumption that no one in their right mind would make, unless a team changed literally everyone from one season to the next, which (afaik) has never happened.

Re: Nice stat of the week - socalstanfan1 - 09-30-2015

(09-30-2015, 12:46 PM)OutsiderFan link Wrote:Stanford used to be a turnover forcing and sack machine.  What the hell happened?

See: condition of the DL

Re: Nice stat of the week - CTcard - 09-30-2015

(09-30-2015, 12:46 PM)OutsiderFan link Wrote:Stanford used to be a turnover forcing and sack machine.  What the hell happened?

Stanford certainly has been a sack machine since the D got good in 2010. Since then Stanford has been in the top 12 nationally in sacks/game every year - usually top 5.

Turnovers, not so much. Stanford has been in the bottom half of the conference in turnovers forced for every year that cfbstats.com has data (that is back to the 2008 season) except for 2010. In 2010, Stanford was #2 in the conference and in a five-way tie for twelfth nationally in turnovers gained.

On turnovers, in most recent years Stanford has been a low-turnovers forced/low turnovers given team.

Re: Nice stat of the week - BostonCard - 10-01-2015

https://twitter.com/TedMillerRK/status/649691033433870337

If turnovers really are random, Cal's defense may have benefited by being lucky.  ANd one wonders what happens when the turnovers regress to the mean.

BC

Re: Nice stat of the week - washingtonismoney - 10-01-2015

Bill Connelly's method of calculating turnovers luck was referenced earlier in this page. According to the 59ers' stat profile (http://www.footballstudyhall.com/pages/2015-california-advanced-statistical-profile), they've been benefitting from a substantial +2.39 points per game in turnover "luck."

(Stanford, by contrast: -5 ppg. http://www.footballstudyhall.com/pages/2015-stanford-advanced-statistical-profile)

Re: Nice stat of the week - fullmetal - 10-05-2015

Perhaps a philosophical shift from the departures of Fangio and Mason?