We need more meaningful stats-Ed Cowan

[ankitj]

I quit reading the article half way through. Looks like I am the only one to do so.

 

[Furball]

I think the most interesting aspect of this article is how he believes this will happen anyway due to organisations like the IPL, where business acumen comes into play.

Should keep an eye on it.

Yeah, it's certainly something the IPL should be using more of. I was actually considering doing a Moneyball style squad before the last IPL based on my criteria, but couldn't be bothered.

 

[Prince EWS]

Awesome article.

The point is you don't necessarily need SABRmetric levels of detail to exist. Any gradual improvement through further analysis would be beneficial.

There has to be some reasonable batting statistic which is able to tell me that Rahul Dravid is much better than Mohammad Yousuf even though Yousuf averages more than Dravid.

Edit: For some reason I thought Yousuf retired with an average of 55-56. Obviously dropped heavily after his epic 2006. Point still stands, replace Dravid with Inzy.

This?


As for the article - yeah, it's natural to read/watch Moneyball and think a similar thing could happen with cricket, but I'm just not sure how realistic it is.

I definitely think more statistics could be collected on fielding, but that's more a case of just starting to record things better than overhauling our existing statistical measures. As for batting and bowling, I certainly think deeper analysis into things like dot ball percentage, boundary percentage, good length percentage, offside/legside breakdown etc would be excellent tools to work with from a coaching or planning standpoint; you could statistically identify where players needed to improve (or where their weaknesses were if you were up against them) and really advance your side. It could also be used to better balance limited overs side with players filling roles more suited to their skills. I think a lot of that already happens to an extent with the England side.

However, in terms of measuring who was better, batting and bowling averages do the hard work for us and take all the different means to an end into account and give us the end result. We can all have our opinions on what attributes matter more in making a good batsman or a good bowler but we can all be disproved if they don't score runs/take wickets. Short of taking into account the standard of the opposition and the pitch conditions, there's not a better measure out there for actual performance and quality than the ones we've got IMO.

EDIT: I would be pretty interested in seeing if there was a correlation between how players scored their runs in domestic cricket and if they were able to transfer their run-scoring to international cricket or not, actually. That's something that could be examined.

 

[Furball]

I certainly think deeper analysis into things like dot ball percentage, boundary percentage, good length percentage, offside/legside breakdown etc would be excellent tools to work with from a coaching or planning standpoint; you could statistically identify where players needed to improve (or where their weaknesses were if you were up against them) and really advance your side

Yeah, this is what I was talking about. I don't think batting strike rates tell enough of a story when it comes to analysing a player, because there's more than one way to skin a cat. There's generally 2 ways a player will end up with a strike rate of around 100 in an ODI - because they're like Kieswetter and are very boundary reliant, or they're like Mike Hussey and don't generally use up a lot of dot balls.

I think Moneyball ideas could be particularly useful in the IPL, because I think there's certain skills that are over-rated in T20 cricket. Chief among them (IMO) is power-hitting. Batting in general is massively over-valued in T20 cricket, but power hitting even more so. Most people can recall innings like Yusuf Pathan slogging a ton for Rajasthan in IPL3, but how many other memorable innings does he have? Powerhitters will very occasionally come off to a massive extent but that sort of innings doesn't happen often enough for franchises to justify spending the money they do on players like Pathan.

I think a lot of that already happens to an extent with the England side.

Matt Prior came back into the ODI side in Australia as an opener.

edit: from the IPL auction thread:

I haven't followed the entire auction in great detail, but I'm going to post a list of my suspicions about what the conventional wisdom is for the IPL and draw conclusions from it, and see how much of my suspicions are true based on what happens in the auction.

1. Batsmen are over-valued in T20
2. Bowlers are under-valued
3. All rounders are massively over-valued
4. Big hitters are massively over-valued
5. Indians, particularly batsmen, are over-valued (and before anyone jumps in on this point, I understand perfectly well the reasons why Indian batsmen have managed to command such huge salaries in this auction)
6. Australians are over valued
7. English players are under valued
8. Foreign spinners are massively under valued.

 

[Prince EWS]

Yeah, this is what I was talking about. I don't think batting strike rates tell enough of a story when it comes to analysing a player, because there's more than one way to skin a cat. There's generally 2 ways a player will end up with a strike rate of around 100 in an ODI - because they're like Kieswetter and are very boundary reliant, or they're like Mike Hussey and don't generally use up a lot of dot balls.

I think Moneyball ideas could be particularly useful in the IPL, because I think there's certain skills that are over-rated in T20 cricket. Chief among them (IMO) is power-hitting. Batting in general is massively over-valued in T20 cricket, but power hitting even more so. Most people can recall innings like Yusuf Pathan slogging a ton for Rajasthan in IPL3, but how many other memorable innings does he have? Powerhitters will very occasionally come off to a massive extent but that sort of innings doesn't happen often enough for franchises to justify spending the money they do on players like Pathan.

I think you're a bit all over the place with this post. As you say, there's more than one way to skin a cat, but a player's strike rate takes both these factors into consideration. It doesn't matter if you score 6 . . or 2 2 2 realistically.

Are you saying that while the "2 2 2" player isn't better than the "6 . ." player, he's not worse either and would attract less bids from other sides, hence being cheaper?

 

[NUFAN]

Are you saying that while the "2 2 2" player isn't better than the "6 . ." player, he's not worse either and would attract less bids from other sides, hence being cheaper?

I'm pretty sure that's exactly what he's saying.

Of course in the IPL '6 . .' is actually better because they get money for hitting DLF maximums.

 

[Furball]

I think you're a bit all over the place with this post. As you say, there's more than one way to skin a cat, but a player's strike rate takes both these factors into consideration. It doesn't matter if you score 6 . . or 2 2 2 realistically.

Are you saying that while the "2 2 2" player isn't better than the "6 . ." player, he's not worse either and would attract less bids from other sides, hence being cheaper?

The strike rate point is in terms of developing team strategy - where do you want your powerhitter to bat? Do you want him opening to take advantage of fielding restrictions, or do you want him at the end to clear the ropes during the slog? Or do you even want a powerhitter when someone adept at running 2s and rotating the strike will score his runs just as quickly with less risk?

I think power-hitters are over-valued in T20 because I think there's a huge misconception about batting in T20 - I've lost count of the number of times I've heard Bumble etc. go on about boundaries in T20 cricket. IMO, T20 batting is about scoring runs off as many deliveries as possible. I've said it a few times, but Michael Hussey IMO is the ideal as far as limited overs batsmen go - he is very quick between the wickets so can steal singles (and his speed also helps with stealing extra runs, ie making 1s into 2s, 2s into 3s), doesn't tend to get bogged down a lot and rotates the strike well, and he also has the ability to clear the ropes when he needs to. Dhoni's the same.

The misconception about T20 batting leads to people thinking the likes of Kieron Pollard and Yusuf Pathan are good batsmen, which affect how much teams pay for them. Dhoni falls into the "over-priced" category because he is Indian; Hussey on the other hand is hugely under-rated and under-valued as a result. IIRC Chennai got him without any other teams bidding for his services - he was an absolute bargain in comparison to the likes of Andrew Symonds and Adam Gilchrist.

 

[vic_orthdox]

Cost benefit analysis for shots would be interesting.

 

[Debris]

I suspect that whatever statistics they come up with will not be much more meaningful that averages. It may be that you can't use statistics to judge a player exept in the broadest sense (ie batsman who averages 55 is better than one who averages 40). Players can change so much from season to season and even game to game that it all becomes a matter of luck.

The methods used in Moneyball were only successful for the one year and they have not had any real success with it since. Good chance that it was a fluke instead of inspiration.

 

[KiWiNiNjA]

The methods used in Moneyball were only successful for the one year and they have not had any real success with it since. Good chance that it was a fluke instead of inspiration.

Nope.

 

[silentstriker]

Moneyball methods are used by every team now.. One team started it.

 

[Debris]

Moneyball methods are used by every team now.. One team started it.

Just because every team uses it does not mean that it is useful. It could just be a fad picked up because the team that used it seemed to have some success. You can look at the Oakland Athletics record in 2002 and it does not seem significantly different from the few years before it. You can argue that they lost their best players for that season but a winning culture had been established at the club. And, in the end, they did not even make the baseball world series.

This sort of thing happens all the time without some magic formula being involved. The perfect example being the Melbourne Storm in the rugby league this year. It may be that it is useful but I don't see any real evidence. The sample size was way too small to be of any use. You really need to set up an experiment where half the teams use it and half don't over several years and see if there is any significant difference in performance, which will never happen. I still don't see where luck has been ruled out.

 

[KiWiNiNjA]

Ummmmm, no.

 

[Debris]

Ummmmm, no.

It might help if you make an actual argument as to why it is valuable. I am prepared to change my mind if anyone can demonstrate sufficiently that this sort of statistical analysis of players will make a difference. Just saying "no" doesn't add anything to the debate.

 

[DingDong]

The strike rate point is in terms of developing team strategy - where do you want your powerhitter to bat? Do you want him opening to take advantage of fielding restrictions, or do you want him at the end to clear the ropes during the slog? Or do you even want a powerhitter when someone adept at running 2s and rotating the strike will score his runs just as quickly with less risk?

I think power-hitters are over-valued in T20 because I think there's a huge misconception about batting in T20 - I've lost count of the number of times I've heard Bumble etc. go on about boundaries in T20 cricket. IMO, T20 batting is about scoring runs off as many deliveries as possible. I've said it a few times, but Michael Hussey IMO is the ideal as far as limited overs batsmen go - he is very quick between the wickets so can steal singles (and his speed also helps with stealing extra runs, ie making 1s into 2s, 2s into 3s), doesn't tend to get bogged down a lot and rotates the strike well, and he also has the ability to clear the ropes when he needs to. Dhoni's the same.

The misconception about T20 batting leads to people thinking the likes of Kieron Pollard and Yusuf Pathan are good batsmen, which affect how much teams pay for them. Dhoni falls into the "over-priced" category because he is Indian; Hussey on the other hand is hugely under-rated and under-valued as a result. IIRC Chennai got him without any other teams bidding for his services - he was an absolute bargain in comparison to the likes of Andrew Symonds and Adam Gilchrist.

not sure what ur point here is but clearly even in the mlb some managers still go for players that the sabermetrics show are no more valuable than an efficient quiet performer even with all the stats they have it's just how the human psyche works. look at the money closers get. i'm sure the cricket coaches also understand that SR is what matters and not the number of sixes but they can't help it just like in every other franchise sport

 

[Furball]

not sure what ur point here is but clearly even in the mlb some managers still go for players that the sabermetrics show are no more valuable than an efficient quiet performer even with all the stats they have it's just how the human psyche works. look at the money closers get. i'm sure the cricket coaches also understand that SR is what matters and not the number of sixes but they can't help it just like in every other franchise sport

That's my point though. IPL teams could save themselves a lot of cash by shunning T20 "stars" like Pollard or Yusuf Pathan and instead buy players who are much better but less flashy.

Pathan is a slogger who can bowl a bit of "off spin." I'm pretty sure he went for around $2m. Dan Vettori and Michael Hussey combined I think went for less than $1m.

 

[Athlai]

There are very, very few players who should be going for double any other player in the IPL. Malinga is one of them.

 

[DingDong]

That's my point though. IPL teams could save themselves a lot of cash by shunning T20 "stars" like Pollard or Yusuf Pathan and instead buy players who are much better but less flashy.

Pathan is a slogger who can bowl a bit of "off spin." I'm pretty sure he went for around $2m. Dan Vettori and Michael Hussey combined I think went for less than $1m.

it's still a new league and a new format and people are still finding their feet im sure eventually they'll settle to some base rules for buying players u'll always have exceptions and that's what makes it interesting
i agree with u that some players are under paid but that's just how it will always be i'm sure if they go back they will no doubt pay a lot more for hussey

disagree on dhoni he's the face of the franchise and probably the most markatable indian player i'd pay the same amount for him honestly it's not always about performance

 

[Furball]

it's still a new league and a new format and people are still finding their feet im sure eventually they'll settle to some base rules for buying players u'll always have exceptions and that's what makes it interesting
i agree with u that some players are under paid but that's just how it will always be i'm sure if they go back they will no doubt pay a lot more for hussey

disagree on dhoni he's the face of the franchise and probably the most markatable indian player i'd pay the same amount for him honestly it's not always about performance

I think Dhoni is over-priced but he's the one Indian player I'd be willing to break the rules for in the IPL.

 

[DingDong]

I think Dhoni is over-priced but he's the one Indian player I'd be willing to break the rules for in the IPL.

free market mate. indians want to see indian stars in their side so they will always get more money plus the whole mandatory 6 indians or whatever rule in the ipl so they will always make more money