Friday 29 June 2018

Using Added Value to measure cricket performances - Part 2 ODI bowling

Reid and Matthews chat before the final over.
It was the summer of 1990/91, just before Christmas. I was on holiday at my Aunty's place in Mount Maunganui. My cousin and I were sleeping in the glass conservatory, looking out over the sand-dunes. The air smelt like salt and sand.

In our little room was a little TV, on the TV was the cricket coming out of Australia.

New Zealand were playing against Australia in Hobart.

I was 11 years old, and I was enthralled.

Danny Morrison was bowling. A year earlier, he had come to speak to my primary school assembly, then signed autographs by the school pavilion. I got him to sign my cricket bat, and I tried really hard to not get it scratched off. He was my favourite bowler. Australia needed 6 runs to win. Greg Matthews was on strike. I wasn't sure why, but as nobody seemed to like Greg Matthews, I didn't either.

Morrison bowled from around the wicket, and speared a fullish ball into leg stump. Matthews drove it, inside-out, through point for four. I really didn't like Matthews now.

It was now down to the last over. Australia needed 2 to win. Bruce Reid was batting against Chris Pringle. Pringle was a bit of a mystery. He had played a couple of matches for Auckland, then got picked for New Zealand against England, despite not being in the tour party. He was just being used as a net bowler, when a couple of players got injured, and he was told "you're playing."

He did well enough, that less than 7 months later he was opening and closing the bowling for NZ. But we still hadn't really learned to trust him.

He had 9 overs, 1/34. Now he was going to be asked to bowl to the human version of the leaning tower of Pisa, to try to defend 1 run.

Bruce Reid was as skinny as he was tall, and he was very, very tall. Greg Matthews had to almost take his helmet off to talk to him, because the visor meant that he couldn't see up that high. I was sitting on the mattress on my bed, but I was finding it difficult to stay seated.

The first ball really should have been called a wide. Steve Randell went against the grain of Australian umpires of his era, and actually made a call that went against Australia. Ian Chappell commented "you'd have to say, probably in any other circumstance that's gotta be a wide, but that'd be a hell of a one to call at that stage."

The next 3 balls were all too good for Reid. He managed to hit one straight to a fielder, but missed the other 2. Pringle was bowling length to him, and it was working.

Now 2 from 2 needed. Matthews wandered down for another chat. I can't imagine what he said, other than "just hit the bloody thing."

Pringle bowled another length ball, this time on about 5th stump. Reid swung mightily over midwicket, missing the ball by about 20 cm.

Last ball. It's now all on Reid. Pringle bowls a delivery that now would be called "hitting the wide hole." At the time it was possibly a misdirected yorker on 5th stump. But Reid had no chance. The ball was in Young's gloves, before Reid had finished swinging the bat. He set off to try to run a bye, But Young was already running in, and he knocked off the bails before Reid had even made it 2m down the wicket. The Kiwis were celebrating, as Bruce Reid walked off, completely bewildered.

Pringle ended with 1/34 off his 10 overs. But his performance was far more memorable than it was statistically impressive. Pringle wasn't even man of the match. That went to Brian Young for scoring 41.

In my last post, I looked at ways of assessing the quality of batting innings statistically by looking at the value added.

I tried applying the same principle to bowling.

I'll look at some of Pringle's most memorable bowling performances and see how they stack up.

Traditionally they are ranked by first the number of wickets, then the number of runs. so 4/88 is considered to be better than 3/21, no matter how counter-intuitive that feels. Here are some selected performances ordered traditionally.

Selected performances of Chris Pringle. 
A couple of things stand out here. Firstly, there are a lot of performances in Australia.

Pringle lived in Australia for about 13 years growing up, and possibly his style suited their pitches more.

Secondly, is the 11 overs bowled in Birmingham in 1994. This was while England were still trying to tinker with the ODI format, and were refusing to play it the same way as everyone else. I'm not entirely sure why they insisted on doing this, but it may explain in part why England have traditionally been the great underachievers in ODI cricket.

Just as you can get a batting index, you can also get a bowling index, that combines runs, wickets and balls. This time its runs squared, divided by the product of the balls bowled and the wickets taken.

Here's how they rank with this format:
Bowling index method for Pringle's innings.

These look somewhat better, but there are some issues.

If someone takes 0 wickets, they get an index of infinity, which seems a bit off.

Another issue is that the difference between 2 and 1 wickets is significantly higher than the difference between 3 and 4 wickets. 3/28 doesn't seem to be 50% better than 5/45, but the index method would rank it roughly that much better.

There's also the fact that there's no context added at all. Madden Lal's 3/15 vs East Africa in 1975 comes out as better than Paul Collingwood's 3/16 against Zimbabwe in 2004, or even Kuldeep Yadav's 4/23 vs South Africa at Cape Town. Not many people would put those 3 performances in that order.

So I decided to add in context. This time I looked at average runs per wicket (ARPW) and average runs per over (ARPO)

Here are the contextual numbers for Pringle's performances:

Average Runs per Wicket and Average Runs per Over expected for selected performances by Chris Pringle
The most interesting thing here is how the expected scores against Australia in Australia changed between 1990 and 1994. Wickets were about the same value, but the runs per over had increased by 0.1; that corresponds to Australia scoring an extra 5 runs per innings

For the added value, I looked at the runs that an average player was expected to conceded, then subtracted the number of runs the player actually conceded.

Here are these innings, sorted by added value.

Chris Pringle's selected performances, sorted by added value.
These look more intuitive. 3/21 off 10 overs feels slightly better than 4/35 off 8.4. There is still no bonus for the fact that he was generally bowling at the death, but again, that's an artifact of the fact that the database that I have is not over-by-over, but rather innings-by-innings.

Using this method, then, what are the greatest performances of all time?

Top 10 Bowling performances
The top result was Ajantha Mendis, just after he broke onto the international scene, and before the batsmen figured out how to play him. Then came Southee's demolition of England in the World Cup. Shane Bond is the only one with 2 entries in the top 10, and there's generally more diversity in the bowling numbers than in the batting ones. In the top 30 there are 25 different names. (Bond is there 3 times, Starc, Boult and Afridi are all there twice)

Finally, the top 10 bowlers by total added value, and average added value:

Most value added (left) and most value added per innings (right)
The average numbers are quite pleasing in terms of the spread. 2 fast bowlers, 2 left arm swing bowlers, two tall right arm metronomic bowlers, and 4 tricky-dicky off spinners. The lack of leggies is surprising, but is not likely to last, given the rise and rise of Rashid Khan, who would be top if I didn't set a limit of 50 matches.

Go to the first article - ODI Batting

No comments:

Post a Comment