First Chance Average?

[Son Of Coco]

It's not impossible to accurately measure said things.

It is impossible to accurately measure some things, and as such I don't try.

Really? You've never given a stuff about it before.

Not trying is not the same as something that's 'impossible'. Impossible to you is possible to someone else. I don't care for the theory, but I can see what you're trying to do. I think the theory itself is massively flawed and you dont have enough of an understanding about what you are trying to measure and what is required when you do that. When presenting statistics you don't simply discard everything you either can't be bothered measuring, or deem it impossible to measure. These variables have to be taken into account and if there are too many then the theory has no statistical basis.

 

[Matt79]

Look, for me the final nail in the coffin of First Chance Averages as a useful measure is that you can't take them at face value. Sure, scorebook averages don't tell you anything like the full story in terms of how they achieved that score, but you know absolutely how the formula is derived. A scorebook score (if that expression makes sense) is a known quantity - it is how many runs a player scored before they were given out by the umpire. There are many nuances not conveyed in that, including whether the decision to dismiss him was correct, but the facts of the matter are clear, ie. he was given out on this score.

First chance scores don't have that certainty. I know that whenever I saw a first chance score that differed from the scorebook score, ie. where some anonymous statistician, inured and insulated from review or accountability for their judgements in a manner very unlike the professional umpires who actually officiate the game, I'd want to review the "chance" myself to be satisfied as to whether the 'chance' was indeed "out". Now, if I'm going to have to do that regularly when reviewing first chance scores, then it's a stat that doesn't offer much of a practical advance in terms of being able to get an impression of what's occurred without watching the innings myself. And given there are some decisions that people divide into two camps over and frequently cannot be persuaded towards an agreement, you're going to end up with stats that are not universally accepted, and which will therefore be good for little than your own private amusement...

 

[FaaipDeOiad]

Top post. Because of the unquantifiable and unprovable nature of the "dismissals" used to determine the stat, you could never simply see a first chance average and make a judgement based upon it, which is after all the purpose of statistics in evaluating performance. Effectively, it demands that you have seen all the games in question to determine just how "lucky" the batsman was, and then re-evaluate the significance of each "dismissal" point to be able to use the stat in any reasonable way. And really, if you've seen all the games, you don't need an incomplete statistical reference point to be able to judge the luck of the batsman. You can quantify that pretty easily in a subjective manner, and indeed incorporate many more things into the judgement than a black and white call about whether or not the batsman was dropped, including the control they had throughout their innings, the conditions it was played in and other things.

 

[Mr Mxyzptlk]

A batsman is out when the opposition is good enough to dismiss them. That includes fielding as well as bowling and any other factors which might come into the equation.

Top post.

And what of batsmen who are out to spectacular bits of cricket (catching, bowling, fielding) from the opposition when they really haven't done much wrong to be dismissed. Does that make them not out in their first-chance average?

Not sure if that's been mentioned before. No time to read back. Essay due in a couple of hours.

 

[GIMH]

Top post.

And what of batsmen who are out to spectacular bits of cricket (catching, bowling, fielding) from the opposition when they really haven't done much wrong to be dismissed. Does that make them not out in their first-chance average?

Not sure if that's been mentioned before. No time to read back. Essay due in a couple of hours.

Excellent point. I wonder if it counts the same as being ran out by an Inzamamesque partner?

 

[Richard]

OK, done. It comes to 30.3. You'll need to do better than a list of numbers. Who is the player so I can compare.

If you have spent so much time with this theory surely you would have a summary of a number of players that you have analysed to prove your point. I would have thought the luck y players such as Gilchrist and Hayden would have been at the top of your list.

Should have been pretty obvious? MET - Marcus Edward Trescothick.

I've not taken down lists of first-chance scores for every single batsman, no, but maybe I'll do one for Sehwag as he's about the only one whose luck has approached Trescothick's over an extended period. Even he, though, has still played some number of good innings.

 

[Richard]

woh, woh, woh....first chance average CREATES more problems than it addresses.

Bull**** does it.

It cannot create problems unless you are of the opinion that batsmen deserve every piece of luck they get.

 

[Richard]

Why is it 'just life' for a bowler, but something that needs to be addressed in a batsman?

Because you can - give me a way to do a similar thing for bowlers?

 

[Richard]

Not trying is not the same as something that's 'impossible'. Impossible to you is possible to someone else. I don't care for the theory, but I can see what you're trying to do. I think the theory itself is massively flawed and you dont have enough of an understanding about what you are trying to measure and what is required when you do that. When presenting statistics you don't simply discard everything you either can't be bothered measuring, or deem it impossible to measure. These variables have to be taken into account and if there are too many then the theory has no statistical basis.

If you want to go that far there are too many variables in anything.

There are things that can't be measured statistically in just about everything - therefore these things have to be discarded when you're trying to make-up a set of statistics.

 

[Richard]

Look, for me the final nail in the coffin of First Chance Averages as a useful measure is that you can't take them at face value. Sure, scorebook averages don't tell you anything like the full story in terms of how they achieved that score, but you know absolutely how the formula is derived. A scorebook score (if that expression makes sense) is a known quantity - it is how many runs a player scored before they were given out by the umpire. There are many nuances not conveyed in that, including whether the decision to dismiss him was correct, but the facts of the matter are clear, ie. he was given out on this score.

First chance scores don't have that certainty. I know that whenever I saw a first chance score that differed from the scorebook score, ie. where some anonymous statistician, inured and insulated from review or accountability for their judgements in a manner very unlike the professional umpires who actually officiate the game, I'd want to review the "chance" myself to be satisfied as to whether the 'chance' was indeed "out". Now, if I'm going to have to do that regularly when reviewing first chance scores, then it's a stat that doesn't offer much of a practical advance in terms of being able to get an impression of what's occurred without watching the innings myself. And given there are some decisions that people divide into two camps over and frequently cannot be persuaded towards an agreement, you're going to end up with stats that are not universally accepted, and which will therefore be good for little than your own private amusement...

So how many stats are universally accepted to mean much? Countless people will tell you that you shouldn't judge cricketers by stats. Mercifully, there is an increasing body who refuses to accept the nonsense that games against Bangladesh and Zimbabwe are the equivalent of games between the top 8.

No statistic is ever going to be close to being perfect. There will be disagreement over a small number of chances, but if people actually took them more seriously, as I've said time and again, it'd be reduced to a minute number. If people discarded the ludicrous term "half-chance" it'd help a lot.

It's not really that hard to tell what a fielder should and shouldn't have caught. Yes, every now and then there will be something indeterminable, and in these instances the simple maxim applies - if doubt, n\o.

There are plenty and plenty of people out there who would be every bit as good at judging what should and shouldn't be out as an Umpire would be at judging what is and isn't out - if they just let themselves, and didn't go down the "half-chance" route.

 

[Richard]

Top post.

And what of batsmen who are out to spectacular bits of cricket (catching, bowling, fielding) from the opposition when they really haven't done much wrong to be dismissed. Does that make them not out in their first-chance average?

Not sure if that's been mentioned before.

It has. My simple answer? That happens.Sometimes you get run-out by a direct hit, sometimes a ball hits the wicketkeeper's gloves and rebounds to the stumps when you're out of your ground - there are all sorts of slightly unlucky dismissals but none come close to the amount of bad luck taken-out when you're given lbw to a ball that pitches outside leg or caught behind to one that flicked your shirt.

For every one of these, there might be a nick past fine-leg, or a top-edge into no-man's-land. These sorts of things happen so often that you never know, they might roughly even each other out.

 

[LongHopCassidy]

These sorts of things happen so often that you never know, they might roughly even each other out.

And in other news, Gelman particularly enjoyed some leftover Christmas ham yesterday before going off to synagogue.

 

[Richard]

That's a blasphemous comment... to 2 religions...

 

[LongHopCassidy]

You get it though, don't you?

 

[pasag]

Hahaha

 

[vic_orthdox]

For every one of these, there might be a nick past fine-leg, or a top-edge into no-man's-land. These sorts of things happen so often that you never know, they might roughly even each other out.

Common consent is that these things even each other out. Anyone who has taken any note whatsoever of the reality of the situation will realise that this is complete and utter bull****.

...

 

[Matt79]

So how many stats are universally accepted to mean much? Countless people will tell you that you shouldn't judge cricketers by stats. Mercifully, there is an increasing body who refuses to accept the nonsense that games against Bangladesh and Zimbabwe are the equivalent of games between the top 8.

No statistic is ever going to be close to being perfect. There will be disagreement over a small number of chances, but if people actually took them more seriously, as I've said time and again, it'd be reduced to a minute number. If people discarded the ludicrous term "half-chance" it'd help a lot.

It's not really that hard to tell what a fielder should and shouldn't have caught. Yes, every now and then there will be something indeterminable, and in these instances the simple maxim applies - if doubt, n\o.

There are plenty and plenty of people out there who would be every bit as good at judging what should and shouldn't be out as an Umpire would be at judging what is and isn't out - if they just let themselves, and didn't go down the "half-chance" route.

"So how many stats are universally accepted as meaning much?" - Um, of the top of my head, some stats that are accepted as meaningful, albeit certainly not perfect, would include: Batting average, batting strike rate, bowling average and bowling strike and economy rates. They require some interpretation and some knowledge to properly be used, but they're basis is clearly known and based on incontravertable facts, ie. what the umpires and offical scorers record at every serious match in accordance with the laws of cricket.

Its utter nonsense to say that "no stat is perfect or universally acceptable" (which is true) but then use that as justification to say that "this deeply flawed proposed system should therefore be acceptable to everyone because no stat is perfect". There are degrees of imperfection, or conversely, degrees of usefulness and soundness for differrent statistics, depending on how they are created.

 

[Richard]

For every one of these, there might be a nick past fine-leg, or a top-edge into no-man's-land. These sorts of things happen so often that you never know, they might roughly even each other out.

Common consent is that these things even each other out. Anyone who has taken any note whatsoever of the reality of the situation will realise that this is complete and utter bull****.

...

Eh? What was the point of those 2 quotes?

I said that one thing does not in any sense even each other out, and I said something else might possibly. So what?

 

[Richard]

"So how many stats are universally accepted as meaning much?" - Um, of the top of my head, some stats that are accepted as meaningful, albeit certainly not perfect, would include: Batting average, batting strike rate, bowling average and bowling strike and economy rates. They require some interpretation and some knowledge to properly be used, but they're basis is clearly known and based on incontravertable facts, ie. what the umpires and offical scorers record at every serious match in accordance with the laws of cricket.

Its utter nonsense to say that "no stat is perfect or universally acceptable" (which is true) but then use that as justification to say that "this deeply flawed proposed system should therefore be acceptable to everyone because no stat is perfect". There are degrees of imperfection, or conversely, degrees of usefulness and soundness for differrent statistics, depending on how they are created.

And as far as I'm concerned the flaws of the two are equal.

Each one has it's problems which give different people problems of taking it seriously. Many people talk disparagingly of stats ITFP.

 

[Richard]

You get it though, don't you?

I get it, I just hope Jews and Christians are not offended by 't.