Definitive proof that not outs don't inflate batting averages

[TheJediBrah]

Agree. To imagine there can be a batsman who scores same number of runs but who gets out on the last ball of each not out innings is where the flaw lies. It's not possible for such a hypothetical batsman to exist because ability to stay not out is causally related to ability to score those runs.

Yeah for someone to believe "not outs boost average" they must genuinely think that when a batsman is not out he is magically more likely to have gotten out soon after and the not out "saved" him or something

I can't even comprehend the logic

Smith averaged 100 plus in the Ashes but I don’t read anything into bcos hurr durr sample size m8

Not sure if srs. Of course you should. No one's suggesting that he'll average 100+ forever, you know why?

sample size m8

dumb post all round

 

[Victor Ian]

....other stuff...Dhoni is probably the only player I can think of that literally bats for his average. ....other stuff....

That's a long wind up to laying the boot into Dhoni (with a side kick to Tendulkar for lols)

Dhoni plays his role superbly. While steering the innings towards victory, he also provides stability at one end to make sure they get there.

Sure, his recent years have been less successful, however just about all players have to run through a late career decline. Bevan got lucky in that he had great players to push him out of the team before his decline. Any fault lies in selection, rather than with the player. When Dhoni finally goes you will realise his value. The replacement might win some games with boom but will also lose a number from being unable to hang around.

 

[Spark]

Dhoni "batting for his average" has somehow managed to coincide with India winning an awful lot of games of cricket with him still there at the end.

 

[stephen]

It was only really in the last year or two that Dhoni started batting for his average in a way that was detrimental to the side.

Bevan was pushed out in 03 so that he didn't have an 07 world cup like Dhoni's '19 world cup.

In some ways Bevan's style of play wasn't one built for longevity since he didn't rely on brute force but finesse and speed between wickets to make his runs. Dhoni was more of a boundary hitter than Bevan did India got a few more years or of him than Australia got from Bevan.

 

[TheJediBrah]

Dhoni plays his role superbly.

Maybe not so much recently though.

 

[Victor Ian]

Maybe not so much recently though.

That would be the, 'Sure, his recent years'...

 

[Prince EWS]

The Johnston was over a season and many innings.

Yeah but the way average works, it was one dismissal. If he got out for a duck in his next innings his running average would go down by half; Smith's wouldn't.

This just shows that when looking at useful sample sizes for batting averages, outs matter more than innings or seasons.

 

[ankitj]

it's poissonian, no? obviously not perfect as there's some correlation between one innings and the next, but i thought that was relatively well-established - at least amongst the people who care about these things.

Seems like Geometric distribution is a special case of Poisson distribution with λ = 1. I think Geometric is right one because it makes 0 the most likely score and makes the frequency distribution continuously falling one which matches with batting scores in real world. Mode for Poisson is λ - 1; set λ to 1 and you get mode of 0.

 

[mr_mister]

i dont think people are really banging on about Smiths average in the recent ashes series that much. i dont remember the number off the top of my head though i know its over 100. averages should never matter that much over one series, season or tournament.


it's more the getting 750+ runs in a series, a truly rare feat. and the fact he did returning from a 12 month ban and while missing a match during the series due to concussion. those things cemented in people's minds as the reasons why it was so special a series.

 

[mr_mister]

Also I know it's classic stat cherry picking, but has anyone ever got six 80+ scores in a row before?

 

[greg]

Also I know it's classic stat cherry picking, but has anyone ever got six 80+ scores in a row before?

Weekes

Records | Test matches | Batting records | Fifties in consecutive innings | ESPNcricinfo.com

 

[Spark]

Seems like Geometric distribution is a special case of Poisson distribution with λ = 1. I think Geometric is right one because it makes 0 the most likely score and makes the frequency distribution continuously falling one which matches with batting scores in real world. Mode for Poisson is λ - 1; set λ to 1 and you get mode of 0.

yeah that makes sense. tbh i don't think it's either but i certainly think a poissonian distribution gives better limiting behaviour i.e. tells you more about how scores are distributed if you ignore the low ones.

edit: i found a paper which suggests that it's negative binomial, which is close enough to poissonian to count for the terminally lazy like me.

 

[mr_mister]

Weekes

Records | Test matches | Batting records | Fifties in consecutive innings | ESPNcricinfo.com

Weekes was a bit of a gun

 

[ankitj]

Actually Geometric is not a special case of Poisson. Setting λ to 1 does make Poisson continuously falling distribution but not really Geometric.

My argument for preferring Geometric (or its continuous equivalent Expinential) over Poisson is that a generalized Poisson can give a non zero score as most frequent score. Looking at Wikipedia, if a batsman averages 50 you have to set λ at 50 for modeling their scores as Poisson. That makes λ-1 = 49 as the most frequent score. That won't happen in real world.

Every batsman's most frequent score is 0, even Bradman's.

 

[ankitj]

Geometric Distribution is special case of negative binomial. I'll give wikipedia reference to not screw it up this time: https://en.m.wikipedia.org/wiki/Negative_binomial_distribution

The geometric distribution (on { 0, 1, 2, 3, ... }) is a special case of the negative binomial distribution, with
Geom(p)= NB(1, 1-p)

 

[Bolo]

Agree. To imagine there can be a batsman who scores same number of runs but who gets out on the last ball of each not out innings is where the flaw lies. It's not possible for such a hypothetical batsman to exist because ability to stay not out is causally related to ability to score those runs.

Yes, but a bat who swings at everything in the closing stages of an odi will average less than one who prods, depite the former being more useful. There may be a way of quantifying it by looking at proportion of not outs in the 1st innings vs not outs in second innings, but I'm not sure if this is reliable, particularly when factoring in unsuccessful chases.

 

[ankitj]

Yeah, different batsmen have different styles hence different averages. And averages are not the only measure of a batsman's worth. Still, not outs don't inflate a batsman's average.

 

[TheJediBrah]

Yes, but a bat who swings at everything in the closing stages of an odi will average less than one who prods, depite the former being more useful. There may be a way of quantifying it by looking at proportion of not outs in the 1st innings vs not outs in second innings, but I'm not sure if this is reliable, particularly when factoring in unsuccessful chases.

That's a very specific, and uncommon, scenario though. It's hardly going to make up a large proportion of individual innings in general, out or not out.

It is the one situation where "not outs boosting average" might have merit. ie. if you're comparing 2 players with the exact difference that you mentioned. I doubt such a situation really exists in reality.

 

[OverratedSanity]

Fmd, someone translate this thread into english for me please

 

[harsh.ag]

Every batsman's most frequent score is 0, even Bradman's.

Wasn't it a big deal when Sachin got his first 0 after a long, long time?