This part of The Plan gasted my considerable flabber.
I mean, really. Someone explain to me how you are accurate to withing 4-5% of whatever it is he’s supposed to be babbling on about. “Oh yes, the computer said Kevin Pietersen would bat like a god at Mumbai, and get out for 188 rather than the 186 he actually managed, and that, yes, the same computer analysis that was 25% out in the selection of the bowling attack at Ahmedabad, was now spot on at Mumbai and Kolkata?” This looks like twaddle.
Someone tell me how this works, because, and I love cricket stats by the way, it is lost on me. “Tell me what Monte” says, said Andy Flower when he couldn’t decide. Maybe whistling out of tune formed part of his statistical analysis.
No wonder we fucked up the World Cup if we had this sort of drivel going on.
A prize to anyone who can decipher that highlighted phrase for me.
Being Outside Cricket 
It means I can play the forthcoming Ashes on International Cricket Captain, send the results to Andy, Trev n Paul and somehow be taken seriously?
And here’s me thinking cricket sims were just a bit of fun.
LikeLike
From wikipedia – on the validity of a given Monte Carlo approach:
– the (pseudo-random) number generator has certain characteristics (e.g., a long “period” before the sequence repeats)
– the (pseudo-random) number generator produces values that pass tests for randomness
– there are enough samples to ensure accurate results
– the proper sampling technique is used
– the algorithm used is valid for what is being modelled
– it simulates the phenomenon in question.
Now Leamon is well educated, so I’m hesitant to say he’s made an obvious error in any of these areas. What I would suspect, paying him due respect, is that actually the sample space of previous Test matches (which is required for calibration, to answer the last two points) is inherently just not big enough to make the approach valid.
As a result, you can keep tweaking the algo to get within 4-5% on historical data – but it’s probably not actually telling you that much with accuracy. Averages work, with some kind of tweak to represent the interaction of the pitch with bowler and batsman talents/preferences.
So why if Leamon is well-educated, might he fall into the trap of thinking that this was working for selections?
Most probably because in fact the number of selection questions put to it was very small. In this period, the key names in the team picked themselves by being head and shoulders above alternatives. So, Monte is actually only being used to think about fringe cases. And when the team was in it’s pomp (or the opposition was in a mess) those fringe cases didn’t really tip the balance one way or the other. So it can all look like it is working, until it properly goes wrong…
Now I’d argue that there were a number of places it went wrong, Ahmedabad, UAE, SA in England, that should really raise big questions – but then I’m a bilious inadequate…
(If you want more in-depth on this kind of thing, someone else may be able to help, or maybe when I’m past June’s work deadline crunch I can come back to it.)
LikeLike
Or maybe it’s pithier to note that at heart maybe the problem was summed up thus: Monte didn’t like Monty… 😉
LikeLiked by 1 person
Aw, man, aw, man, aw, man.
Overfitting!
Nobody should be allowed to mess with any such data analysis software without being told what overfitting is!
LikeLike
It was the players who revolutionised batting and invented all these weird shots. It wasn’t any coach or statistician. I hope England’s performance in the WC has put an end to this nonsense. (It was rather fun to watch though…)
LikeLike
I guess the idea of taking three tall fast bowlers to Aust in 2013/14 was also a part of that statistics based planning. That went well too, didn’t it!
LikeLike
The bar was set low many years ago by Lord Haw Haw Nicholas, who, when England played Australia at the MCG in 1994/5 proclaimed that Ray Illingworth was the best reader of a cricket pitch in the world game. He bought into the nonsense that because El Ill had turned up, things would change for the better.
We got mashed up.
Precedents aren’t just the leader of the United States.
LikeLiked by 1 person
Andrew Strauss has a degree in Economics. This is a fact that should not be forgotten when considering England’s past espousal of number-crunching approaches.
LikeLike
How his book won Wisden Cricket Book of the Year a couple of years ago I will never know. Insipid.
Lest people think I’m a Strauss “hater”, I will go to my grave saying he played the greatest unheralded innings I’ve seen from an England player. His century on the opening day of the 5th Test in 2005 was magnificent. Sure, I love KP’s 158, but that doesn’t happen without the Strauss first innings.
LikeLike
Strauss was a far freer player in his early years. that series in South Africa 2004/5 especially. I was also in Chennai when he got those two tons in the test where it all fell apart. Two fine tons, no matter the final result.
However he was never the same player after the 2009 home Ashes. He got sorted out big time in SA in 2009/10 and never truly recovered. His batting record from that series onward was pretty poor and was overlooked because of the success of the team for a two year period. I suppose I’m speaking on the same hymn sheet as many here when I say it is forgotten too easily that the last few months of his playing/captaincy career were pretty disastrous and that the lack of flexibility from him and Flower cost England dear not only then but as time moved on.
Perhaps it will sound reactive but I remember recalling the UAE Pakistan series and thinking before it started, that Strauss and Morgan were nigh on guaranteed to fail and that if a couple more England batsmen had poor series that the team would really struggle. I was only wrong in so much that every batsman failed!
LikeLike
PKTroll, do you ever hear from the press pack or their assorted BTL fanboys that Strauss averaged 32.8 in the last third of his career? That’s the last 33 Tests, from after the 2009 Ashes. In that period he averaged 60+ against Bangladesh and WI so his record elsewhere (especially SA, SL and Pakistan) is horrendous (21, 17 and 25 respectively).
I wouldn’t mind so much but the same types who ignore the decline of Strauss love banging on about Pietersen’s “career of two halves”. You know, that second part of his career when he only averaged in the low forties rather than 50+ and only played three all-time great innings. The bastard. Just how detached was he? Should have dumped him sooner!
I also wouldn’t mind so much if Strauss wasn’t so sanctimonious about selflessness and “the team must always come first”. Retiring after your hundredth Test and with your average still just above 40? Coincidence? Yet when another SA-born batsman says he wants 10000 career runs it becomes the ultimate crime.
It’s the double standards that stink.
LikeLiked by 1 person
The double standards about Strauss and KP are vomit making. But they do point out with laser like accuracy the hypocrisy of our pro ECB media. And reveal them to be the biased operators and snake oil salesman.
LikeLike
And similarly, it took them until the Caribbean tour to notice Trott’s far more precipitous statistical decline. That was two and a half years after the point at which he’d already played as many games averaging 35 as he did averaging over 60.
The use of Pietersen’s “career of two halves” to belittle him is very high on my list of major irritations, for a multitude of reasons. Not least because even in “decline” his average is over 44, and in between Boycott’s retirement and the huge run-getters of Flower’s side only Gower and Thorpe had managed to sustain that over a career.
LikeLike
SimonH:
I also wouldn’t mind so much if Strauss wasn’t so sanctimonious about selflessness and “the team must always come first”. Retiring after your hundredth Test and with your average still just above 40? Coincidence?
Actually I think Strauss is quite open about that, I have heard him say that he quit because he’d achieved everything he wanted to as a player.
That was when it was being suggested that he left because nasty Kevin was nasty to him, but he said very firmly that no, he’d had certain personal goals and had reached them and then felt it was ‘time to go’. It seemed to me that he’d felt himself going past his peak ability but had decided he wanted to play 100 tests and hung on for that.
I expect he dressed it up as ‘once you feel like that you’re not doing your best for the team’ but I filtered that bit out.
LikeLike
And his nickname was “numbers”
He sounds like a character out of an Ealing comedy. Like Soapy Stephens It beggars belief. But then it doesn’t really. It’s s everything we thought. Anyway, why they keep on about being number 1 is beyond me. Swann says the the number 1 ranking was bullshit.
“”We were No. 1 in the world but if anyone says ‘that’s because you were the best team’ then that’s absolute bollocks. We were nowhere near the best one-day team in the world. We got extremely lucky. We had 18 home games, we had a series against India where, if we struggled, it rained. And it made people think what we were doing was right, hence the stifled approach up until three weeks ago.”
Someone should tell “numbers” Then again someone should tell Flower and Strauss. I wonder if Swann will be accused of destroying the memories like KP was accused of?
LikeLike
I’m not buying Swann as the sinner wot repented like. 18 months ago he was telling us that it would be stupid to let Flower’s knowledge go to waste. Now he’s basically calling Flower, Giles and Moores morons.
We were a daft collapse from winning the Champions Trophy in 2013. We got wickets, and we blew a simple chase. Naff all to do with numbers and stats, and everything to do with poise and bottle under pressure (or lack of it). This is the cricketing equivalent of us playing Spain to a 3-3 draw in football with 10 minutes to go. In a friendly.
LikeLike
I just find it amusing how Swann is allowed to piss on their parade, and the usual suspects nod sagely. But if KP points out the problems the media goes into hysteria.
I always liked Swann,,and his role in the team was under played in my view. Both Staruss and Cook have taken way too much credit instead of Swann at his peak. But then Swann was just another of the poor bloody infantry men. Not a blue blood officer class like Strauss and Mr Waitrose.
LikeLiked by 1 person
This number3 bloke? Didn’t Nick Hoult report that they’d binned him for the ODI’s……I wonder if he saw that coming? A bit like Mystic Meg not foreseeing the demise of the NotW?
LikeLike
OK England’s Statto has a computer model. So do the Met Office and sometimes they get the forecast right, sometimes they don’t. Would any of these number crunchers stake their house on their next prediction?
LikeLike
I would feel happier if John von Neumann (said by loads of people, oncluding Richard Feynmann, to be the most intelligent person they had ever met) were running the simulations rather than someone who got a good first degree from Cambridge. I seriously doubt that there is sufficient data to cover all the variables – pitches, bowling, form, umpiring ………. – to do a meaningful set of simulations. It is like using the Deloitte’s rankings to predetermine the outcome of a match, as if everyone always plays according to the rankings. The caveats/variances around the results must mean that you would have to run millions of simulations to get anywhere close to reality.
How many successtul horse race gamblers have used this approach? And they are not looking at a contest that lasts 5 days. Sheer and utter insanity. Blinded by science. Flowers reveals himself to be a world-class moron with this.
LikeLike
John Bronowski, who presented The Ascent of Man, also thought John von Neumann was the most intelligent person he ever met.
LikeLike