NFL and MLB angles to look at

[jolmscheid]

Hi all...I have been wondering for awhile if, say for instance, that an NFL team comes off a game with a poor performance where they commited 3+ turnovers or something, would it be smart to bet on them the next game as they will probably look to definitely tone down on their turnovers?

The same with MLB? Like if they commit 2+ errors or something?
Same with NBA? 15+ turnovers or something??

What do you guys think? I obviously don't know the past stats of these angles....

[jolmscheid]

The above turnover / error numbers were just examples and NOT set numbers..it's open for variation...basically i'm wondering if a team in any sport is coming off of a bad game of committing numerous turnovers / errors is good to bet ON in their NEXT game...

[Peep]

I don't think so.

I could just as easily make a argument that the team is error or turnover prone.

And one of the best ways to make an error is to try and not make an error.

[LT Profits]

I don't think so.

I could just as easily make a argument that the team is error or turnover prone.

And one of the best ways to make an error is to try and not make an error.

The flip side may be that turnovers are mostly random and the team that committed the TOs may be undervalued next game.

[LT Profits]

I was just referring to football, turnovers may be more predictable in the other sports.

[Peep]

I was just referring to football, turnovers may be more predictable in the other sports.

I don't know if they are "undervalued" next game or not. I don't have turnovers in my DB, so I have no way of checking.

So just expressing an opinion. I could see it going either way, or having no effect whatsoever.

[Justin7]

A turnover changes the outcome of an NFL game by about 4 points on average. Turnovers are roughly 80% luck, so you can do adjustments to power rankings based on turnovers, and get quite a bit more power.

[maxdalury]

You could easily test this trend by getting your own database to check this out. I could probably help you get a db to check this out for pretty cheap if you PM me.

[RickySteve]

I'd be surprised if errors in baseball are more than barely positively correlated to winning %.

[IrishTim]

I'd be surprised if errors in baseball are more than barely positively correlated to winning %.

According to either Bill James or "The Book" (can't remember which one) an error is worth about half a run.

[Igetp2s]

In football, I would focus on fumbles specifically rather than INT's. The fumble recovery rate should be approximately 50%. If a team is recovering significantly more than 50% either of their own fumbles or their opponents' I would bet against them as their luck will even out over time and they may be overvalued.

[skrtelfan]

I'd be surprised if errors in baseball are more than barely positively correlated to winning %.

Me too. Tiny, tiny sample size but last year there was a significant negative correlation between errors and winning percentage. The leaders in fewest errors last year were the Pirates 73, Phillies 76, Twins 76, Blue Jays 76, Astros 78--3 bad teams and 2 playoff teams.

[scratbandit]

Very interesting indeed..

[donjuan]

According to either Bill James or "The Book" (can't remember which one) an error is worth about half a run.

The error itself may be worth half a run but players with good fielding range are going to put themselves in a position to make more errors while at the same time making more outs due to their increased range.

[RickySteve]

The distinction between hit and error is completely arbitrary.

[Justin7]

The distinction between hit and error is completely arbitrary.

So how much is a non-home run hit worth? About the same as an error?

[IrishTim]

So how much is a non-home run hit worth? About the same as an error?

From the same source:

Home run = 1.397 runs
Triple = 1.070 runs
Double = .776 runs
Error = .508 runs
Single = .475 runs

[skrtelfan]

Right, an error should be worth fractionally more than a single because some errors are multi-base errors. That an error is worth about half a run doesn't mean there's little correlation between errors and winning percentage, though. The average team probably faces what, maybe 30 balls in play per game? That's almost 5000 balls in play over the course of a season, and the difference between the team with the most errors and the team with the least errors in MLB last year was only 70. Then factor in that the difference between an error and a hit is a judgment call, and some of those errors would have been turned into an out by very few players, and we're dealing with very subtle differences.

Even the arbitrary nature of errors aside, even if you had an omniscient official scorer who always knew what was an error and what wasn't, they'd still be pretty useless because as defined, they ignore mental errors and also punish fielders with better range, because the fielder with better range is in position to get to a tough ball that might be deemed an error or a hit whereas the fielder with poor range would never get close to that ball. Given two identical teams with identical skills except that the second team has better range, the second team will surely make more errors in the course of the season.

[SparJMU]

I heard an interview one time about the 2007 Ravens and how they experienced terrible luck at recovering fumbles. Therefore this person was predicting a much better record in 2008, which was Flacco's rookie year when no one in the world though the Ravens would be any good. Turns out they shattered thei season O/U and lost in the AFC championship game.

[Peep]

I heard an interview one time about the 2007 Ravens and how they experienced terrible luck at recovering fumbles. Therefore this person was predicting a much better record in 2008, which was Flacco's rookie year when no one in the world though the Ravens would be any good. Turns out they shattered thei season O/U and lost in the AFC championship game.

OK, good example.

Anyone know what teams had the highest and lowest fumble recover rates last year and their number of wins?

We can then look back on this thread in a year and see how it turns out for improvement and falling back in the standings.

[Igetp2s]

I was thinking of more like a game to game change in luck rather than a season by season change.

A team that looked really good one week, but won easily because they recovered 4 of the 5 fumbles in the game, is a team I would rather bet against the following week, all else being equal.

[Wrecktangle]

One of the best single predictors I've used in the past was a TO running average over the last 3 games. As I recall when it got to 6 TOs take the team that had lost them and you got a performance above 55% ATS. It seems that the line overreacted to the teams performance and since the TOs are mostly random, you achieved an edge.