Unified Power Ratings for ODI batsmen at CWC 2015

[kiwiviktor81]

This article introduces a concept called Unified Power Ratings for batsmen in the 2015 Cricket World Cup. This is a way of comparing the effectiveness of the batsmen in the various teams.

The logic of the Unified Power Rating is simple. It is assumed that, in ODI cricket, a batsman's strike rate is as important for his contribution to the team as his batting average. This is because a batsman can strengthen his teams position in two ways (by scoring runs and by scoring them quickly) and a team can lose in two ways (by getting bowled out or by playing out their 50 overs at too slow a strike rate).

Because a good average tends to be roughly half of a good strike rate, the player's batting average is doubled and then, being of equal value to the strike rate, multiplied by it to arrive at a final figure.

The table below ranks the top 40 batsmen in the squads for the 2015 Cricket World Cup by their UPR, or Unified Power Rating (minimum 25 matches).

1. James Faulkner AUS 10692.26
2. AB de Villiers SAF 9888
3. Hashim Amla SAF 9391.59
4. Virat Kohli IND 9372.67
5. Corey Anderson NZL 9355.62
6. MS Dhoni IND 9346.78
7. Luke Ronchi NZL 8676.32
8. Shikhar Dhawan IND 7898.51
9. Kane Williamson NZL 7660.66
10. Quinton de Kock SAF 7496.67
11. Shane Watson AUS 7318.7
12. George Bailey AUS 7237.72
13. Andre Russell WIN 7228.42
14. Glenn Maxwell AUS 7103.17
15. Michael Clarke AUS 7051.99
16. Ross Taylor NZL 6936.75
17. Angelo Mathews SRL 6914.59
18. Jos Buttler ENG 6889.01
19. Paul Stirling IRE 6698.84
20. David Miller SAF 6654.33
21. Joe Root ENG 6649.28
22. Suresh Raina IND 6643.62
23. Tillakaratne Dilshan SRL 6609.18
24. JP Duminy SAF 6535.22
25. Aaron Finch AUS 6396.04
26. Rohit Sharma IND 6348.48
27. Misbah-ul-Haq PAK 6304.58
28. Kumar Sangakkara SRL 6296.81
29. Eoin Morgan ENG 6265.12
30. Ambati Rayudu IND 6231.77
31. Chris Gayle WIN 6205.54
32. Umar Akmal PAK 6096.52
33. Martin Guptill NZL 5937.6
34. Ian Bell ENG 5579.36
35. Brendon McCullum NZL 5551.52
36. Shahid Afridi PAK 5463.77
37. Craig Ervine ZIM 5454.54
38. Brad Haddin AUS 5128.86
39. Grant Elliott NZL 5079.15
40. Ahmed Shehzad PAK 5012.64

James Faulkner, with an average of 48 and a strong strike rate, tops the list, with the South African duo of AB de Villers and Hashim Amla taking the next two positions. Virat Kohli and MS Dhoni (who both average over 50) and Corey Anderson, who averages a solid 36 with a strike rate of 129, are the next three in the list.

New Zealand has three batsmen in the top 10 of this list. Corey Anderson and Luke Ronchi both average in the mid 30s with strike rates above 120, and Kane Williamson who averages 46. India also has three players in this list, with Shikhar Dhawan in at No. 8.

South Africa also has three batsman in the top 10, with Quinton de Kock in 10th place.

Paul Stirling is the highest ranked batsman from the minor nations, in at 19.

Australia, despite only having one player in the top 10, has 5 in the top 20.

Perhaps in discordance with his current form, Brendon McCullum doesn't feature until No. 35, lower than his opening partner Martin Guptill (33). This is because of McCullum's relatively low average of a shade over 30.

Original here.

 

[marc71178]

Of all the statistical manipulations ive seen this is pretty near the worst.

 

[cnerd123]

Heh. I think it's not bad. The list looks pretty good too. It's far from 100% and the results have to be taken with a grain of salt, but hard to argue against those 40 names being more or less the 40 best ODI batsmen at the World Cup.

 

[Riggins]

Yeah it actually gives a surprisingly good result.

 

[Dan]

Yeah, in terms of coming up with 'best batsmen', it's not useful. But certainly as a 'most valuable player' kind of thing it isn't too far off -- to me, 8 of the top 10 guys there are guys you would love to have in your team. Forkers so valuable with the bat these days finishing off the innings.

Strong bias towards power-hitters though, but given the rankings are all about power, that's kinda the point.

 

[viriya]

Because a good average tends to be roughly half of a good strike rate, the player's batting average is doubled and then, being of equal value to the strike rate, multiplied by it to arrive at a final figure.[/URL]

This doesn't really make sense. Doubling the average and then multiplying by SR will not change the order of the players - just the numbers. It's the same as just doing average * SR. To give more importance to average you would need to subtract a constant from it before multiplying by SR, so:

(average - x) * SR
where X is something like 10.

instead of

(average * 2) * SR which is just 2*(average*SR).

 

[hendrix]

Doubling the average and then multiplying by SR will not change the order of the players

uh, yes it will...

 

[Riggins]

Nah it wont.

 

[Prince EWS]

uh, yes it will...

The doubling is pointless.

If 2x * y = z, then x * y = z/2. The numbers change but the order of the players wouldn't at all.

 

[hendrix]

you're right, I mis-understood what he was saying.

 

[hendrix]

I thought it was average multiplied by 2 and then added to the SR

i.e. player x average 5 SR 10 = 20

player y average 7 SR 8 = 22

 

[Days of Grace]

How much weight do you give to average vs. strike-rate for ODI batsmen? 60/40 in favor of average?

On a related note, I've been thinking of how to weight strike-rates and economy-rates for bowlers in both tests and ODIs. Does 60/40 in favor of strike-rates in tests and vice-versa in ODIs sound reasonable?

 

[viriya]

How much weight do you give to average vs. strike-rate for ODI batsmen? 60/40 in favor of average?

On a related note, I've been thinking of how to weight strike-rates and economy-rates for bowlers in both tests and ODIs. Does 60/40 in favor of strike-rates in tests and vice-versa in ODIs sound reasonable?

Should be higher imo. Econ barely matters in Tests and while SR is nice in ODIs, unless you're taking quality wickets Econ is much more important. Especially since low SR ODI bowlers are usually pacers who tended to take cheap wickets at the end of the innings. Waqar is a good example - he has a lot of 5-wkt hauls but not all of them are what you would call "quality".

 

[kiwiviktor81]

This article assumes a 50/50 split between strike rates and averages. It's an interesting experiment, because you can judge batsmen on all manner of factors.

The point of doubling the average was to use that sub-factor as a comparison to strike rates. I.e. if a batsman has a strike rate of 90, we can say that this is the equivalent of an average of 45. This might be unrealistic when looking at batsmen like Corey Anderson, whose strike rate of 129 would translate to an average of 64.5.

In my opinion the relatively high value put on averages is a remnant from the Test match culture that existed before ODIs and which still influences it. In the same way that 230/4 after 50 overs was once seen as a winning score, but now is seen as a waste of lower-order batting resources.