First Chance Average Revisited

[Fusion]

I know Richard hasn’t posted in ages, but let's keep out the personal insults please while discussing this...hallowed topic.

 

[Chubb]

Part of the problem with FCA was that RIchard did not rate many batsmen at all. He is a bowler who genuinely - and passionately - believed "the batsman can never be too good for the bowler". So his formulation was biased against batsmen in general because he believed, simply, cricket was a batsman's game.

 

[Howe_zat]

baseball stats take it into account don't they? tracking fielding errors, and as an example not counting things as a hit if it was due to fielding error.

That's why this might make some sense if it included all chances*, but it doesn't, it's only the first. By this method an innings where someone is dropped on 0 and then 3 times on their way to 37 is rated the same as someone who's dropped on 0 and then goes on to get 350. Which is obviously crocked.

*and assuming a "chance" is remotely quantifiable etc etc, which is the real issue

 

[morgieb]

I prefer the Fat Chance Average personally.

 

[Riggins]

*and assuming a "chance" is remotely quantifiable etc etc, which is the real issue

yeah, i don't really have any knowledge about baseball, but wouldn't error determination be just as subjective?

 

[Hurricane]

Just to play devils advocate. I remember thinking about the merit of first chance averages when Ponting had a dolly dropped against pak at fine leg and went on to make a career prolonging double.

8.4
Mohammad Asif to Ponting, 1 run, dropped, short ball, hooked high out to the deep, straight down the throat of Aamer at long leg who puts down a sitter

 

[NUFAN]

He made 197, just sayin'.

I wouldn't mind a balls faced per chance stat tbh.

 

[Hurricane]

3rd Test: Australia v Pakistan at Hobart, Jan 14-18, 2010 | Cricket Scorecard | ESPN Cricinfo 209

 

[NUFAN]

Hurricane 1 - NUFAN 0.

 

[Howe_zat]

yeah, i don't really have any knowledge about baseball, but wouldn't error determination be just as subjective?

I don't know how it's done in basball either. But I'm guessing, at the very least, you get to avoid the much harder to pigeonhole nearly-outs you get in cricket like lbw appeals, missed stumpings and poor captaincy meaning the fielder's just not there to take it.

 

[fredfertang]

Fred you are a top bloke, so I say this advisedly:

THIS ENTIRE ****ING IDEA WAS A HORRID BLIGHT ON CRICKETWEB FOR YEARS, HAD NO BASICS IN FACT OR REMOTE COMMON-SENSE. IT WAS INVENTED FOR SOMEONE TO TALK DOWN PLAYERS HE DIDN'T RATE. TO EVEN MENTION IT AGAIN IS ****ING ********.

and yes, I am shouting.

I thought the fact that someone had done the research for Bradman was interesting and justified a mention, not because it in any way legitimises the FCA, but because of what it says about Bradman's superiority

Trouble is there was no way of making the thread without raising the spectre of Mr Dickinson and I had hoped my tackling the issue in the way I did would steer any conversation on to what the stat said about Bradman rather than what it said about the FCA and my reference to ODIs was intended as the clue that my tongue was planted firmly in my cheek

But it seems that Richard's ghost haunts this place still, which is a shame but 'appen I was wrong to kick this off, and perhaps the thread ought to be locked

....................... anyone got a Ouija board?

 

[NUFAN]

I don't know how it's done in basball either. But I'm guessing, at the very least, you get to avoid the much harder to pigeonhole nearly-outs you get in cricket like lbw appeals, missed stumpings and poor captaincy meaning the fielder's just not there to take it.

It is subjective in Baseball, commentators frequently debate that the hitter was unlucky for it to be judged an error or visa versa. Its just that over the course of a season or career a couple of debatable errors or outs don't make much of a deal at all.

 

[Furball]

I thought the fact that someone had done the research for Bradman was interesting and justified a mention, not because it in any way legitimises the FCA, but because of what it says about Bradman's superiority

Trouble is there was no way of making the thread without raising the spectre of Mr Dickinson and I had hoped my tackling the issue in the way I did would steer any conversation on to what the stat said about Bradman rather than what it said about the FCA and my reference to ODIs was intended as the clue that my tongue was planted firmly in my cheek

But it seems that Richard's ghost haunts this place still, which is a shame but 'appen I was wrong to kick this off, and perhaps the thread ought to be locked

....................... anyone got a Ouija board?

Bradman's batting average says enough about his superiority and you still get muppets questioning it.

 

[flibbertyjibber]

I prefer the Fat Chance Average personally.

Cosgrove wins that one now Leverock has retired.

 

[Top_Cat]

SABRmetrics isn't just about errors. There's some subjectivity about it but it's generally about the analysis of easy to collect, objective data like RBI's, bases stolen, etc. It then uses these to test hypotheses.

I don't like it that much as an analysis method but, compared to cricket statistical analysis, if SABR is a scientific calculator, cricket is cavemen counting rocks in the dirt.

 

[BoyBrumby]

I'm probably giving the idea way more credence that it deserves by giving an even semi-serious reply, but the main problem I saw with FCA is that it takes no account of shots that the batsman have no control over. The edge that flies at a catchable height in the gap between 2nd slip and gully, say.

It's a batsman's error, but one for which he gets (in all likelihood) four runs for instead of a "dismissal" against his name if exactly the same shot had gone to third slip's hands and then been Matthew Waded.

 

[Manee]

I think it is not useful for comparing batsmen who have had completed careers. It is only useful for assessing whether a young batsmen is going to have a full and successful career. A couple chancy hundreds are not as promising as two chanceless ones. But then this operates over a short time period so the 'FCA' as an average isn't needed at all. I do miss Richard though, his views were always intriguing.

 

[Top_Cat]

The massive problem with the internal validity of a first-chance average is that the logic which seems to only apply to catches can and should apply to all modes of dismissal. Because a batsman can be out in more ways than merely caught, you have in-built bias in any calculations you do from there. This is aside from the actual definition of what consitutes a 'chance' which can and does introduce more bias. For example, Richard's hard limit was if the ball went to hand but I doubt anyone would accept that as anywhere near an adequate definition of a chance for fairly obvious reasons.

It fails at any level of logical investigation if you want to make it generalisable outside of a batsman's potential score when only taking into account catches which go to hand rendering it pretty much a pointless exercise.

 

[fredfertang]

I'm probably giving the idea way more credence that it deserves by giving an even semi-serious reply, but the main problem I saw with FCA is that it takes no account of shots that the batsman have no control over. The edge that flies at a catchable height in the gap between 2nd slip and gully, say.

It's a batsman's error, but one for which he gets (in all likelihood) four runs for instead of a "dismissal" against his name if exactly the same shot had gone to third slip's hands and then been Matthew Waded.

Quite - on that basis, and in all seriousness, Graeme Fowler's FCA would be lower than Chris Martin's - and Clive Lloyd's would be none too flash

 

[Hurricane]

Did Richard go to the extent of calculating them (FCAs) for players?