Cribbage's Standardised Test Averages (UPDATED November 2018 - posts 753-755)

[Flem274*]

Edit - People who missed out because of not playing tests because their team didn't play tests in those years should miss out, what value does a cricket player have if the team isn't playing cricket.

I wouldn't call what Australia played in the 4th test cricket but I bet they'd kill puppies to put Smith in that team again.

 

[Athlai]

Don't have a Cowie man.

 

[NUFAN]

I haven't focused on the bowling list properly but the batting list is very interesting.

One of the first players I looked at that seems low is Gilchrist down in 113th place. I can understand why though as part of his appeal was his ability to turn a match in a session and being behind a quality batting order it doesn't help either, plus he didn't finish that well.

Siddle down in 929th I think you need to implement a new criteria which includes 2 50s in the same match although that might bump up Root too high.

 

[Prince EWS]

Right, so the following players have apparently been affected by what I'll call the "Headley bug" - that is, due to a quirk in the way it estimates how much players would have played in years their countries played zero Tests within the bounds of their career, playing an extra game at some point has actually made their longevity go down.

                        Max     End

Majid Khan (Pak)	13.57	13.55
SN McGregor (NZ)	6.61	6.60
RS Modi (India)	        3.11	3.11
MH Mankad (India)	10.11	10.09
CJ Barnett (Eng)	3.90	3.89
Intikhab Alam (Pak)	13.44	13.39
G Kishenchand (India)	1.53	1.53
HR Adhikari (India)	5.43	5.41
KG Viljoen (SA)	        10.44	10.40
FE Woolley (Eng)	15.66	15.58
GA Rowe (SA)	        3.27	3.25
Javed Burki (Pak)	7.50	7.42
PGZ Harris (NZ)	        2.64	2.61
AE Hall (SA)	        2.10	2.07
JW Hitch (Eng)	        2.28	2.24
Nasim-ul-Ghani (Pak)	7.70	7.57
ER Mayne (Aus)	        1.13	1.10
JWHT Douglas (Eng)	7.46	7.26
FW Smith (SA)	        2.67	2.58
RK Nunes (WI)	        1.50	1.45
P Roy (India)	        8.10	7.78
CP Mead (Eng)	        6.76	6.33
WM Wallace (NZ)	        10.00	9.32
GA Faulkner (SA)	8.31	7.58
PHJ Trimborn (SA)	2.00	1.82
Hon.LH Tennyson (Eng)	4.50	4.00
CMH Hathorn (SA)	5.50	4.76
JJ Kotze (SA)	        1.67	1.41
A Melville (SA)	        6.50	5.46
M Bisset (SA)	        2.00	1.67
CG Fichardt (SA)	2.50	2.08
S Mushtaq Ali (India)	7.04	5.81
CS Nayudu (India)	7.33	5.65
VM Merchant (India)	9.00	6.83
PAM Hands (SA)	        5.00	3.70
GA Headley (WI)	        11.50	8.40
WH Copson (Eng)	        1.33	0.89

Most of them are really slight deviations to the point where I think some of them might even just be rounding errors, but that's a fair list of players so I'll incorporate a fix. For now I'm just going to let them have their "max longevity" that they achieved earlier in their careers (I'll go and edit posts 697 and 698 as they're too minor changes to make new posts for; Headley does jump to third though), but I'll try to come up with something more elegant for the next major overhaul, whenever that is.

 

[Prince EWS]

For those who are interested in what it'd look like just giving no credit at all for years in which teams played no Tests (such as weldone), I've created a Google Doc for that.

https://docs.google.com/spreadsheets/d/1Dulfao4Zy_gwmHePVRvBD6MsGAw2m4C4R7Eddnuw3Uk/edit?usp=sharing

Cowie moves down to 22nd, Headley is 12th. The results probably 'feel' more right like this, but I don't really philosophically agree with the methodology so I won't be making this the official change. I'll try to remember to update that doc with the alternate calculation for the people interested in it though, whenever I do an update to the main one.

 

[ankitj]

Standardized averages of Steyn and Philander are astonishing. Probably shows we underrate modern bowlers because of this being a generally batting friendly era leading to relatively higher averages.

 

[OverratedSanity]

Very surprised at Morkel's standardised average being that low.

Also how did Bond, Harris and Shoaib end up with such similar overall value despite Shoaib playing way longer.

 

[Prince EWS]

Very surprised at Morkel's standardised average being that low.

Also how did Bond, Harris and Shoaib end up with such similar overall value despite Shoaib playing way longer.

Their standardised averages are a bit better (20 to 23 is a 15% difference, afterall), and Shoaib missed so much cricket for Pakistan throughout his career that his "way longer" only ended up being worth two years or so. It square roots the longevity when working out the "value" so a longevity of 2.5 only ends up being worth about 25% less than a longevity of 4.5; I think this reflects how we (or for me, at least) tend to view careers fairly well. He still ended up slightly above them, too.

 

[akilana]

Wake up
Open thread
Check Kallis’s rating
Approve method used
Go to bed.

 

[vcs]

Fantastic work, PEWS.

 

[Dendarii]

How does the longevity calculation work for players who played for more than one country. I was thinking specifically of John Traicos, although he didn't play too many tests. So while perhaps it doesn't matter too much for him, Kepler Wessels did play a decent number of matches for both his sides.

 

[Prince EWS]

How does the longevity calculation work for players who played for more than one country. I was thinking specifically of John Traicos, although he didn't play too many tests. So while perhaps it doesn't matter too much for him, Kepler Wessels did play a decent number of matches for both his sides.

Calculates them as separate players and then just adds them together at the end.

 

[Flem274*]

current unretired top 10 bowlers as dictated by pews

1 - steyn
2 - philander
3 - anderson
4 - broad
5 - rabada
6 - herath
7 - ashwin
8 - boult
9 - jadeja
10 - southee

and the batsmen

1 - smith
2 - ABDV
3 - kohli
4 - amla
5 - cook
6 - root
7 - pujara
8 - williamson
9 - taylor
10 - warner

so basically pews rankings think australian bowlers are overrated crap and so is big daddy kane who no longer belongs in the fab 3

 

[Spark]

Wake up
Open thread
Check Kallis’s rating
Approve method used
Go to bed.

If Kallis isn't rated highly by a Cribb statistical methodology, then you can assume that his account has been hacked

 

[Prince EWS]

so basically pews rankings think australian bowlers are overrated crap

Haha they basically just think they've not played a lot yet tbh. The ratings aren't really designed to measure current ability; just entire careers.

Standardised averages might be more useful than rankings for players who haven't retired yet. If you just use those you get:

Batting:
1. SPD Smith (Aus) - 61.25
2. V Kohli (Ind) - 53.52
3. JE Root (Eng) - 52.91
4. CA Pujara (Ind) - 51.07
5. AK Markram (SA) - 50.04
6. AB de Villiers (SA) - 49.18
7. DA Warner (Aus) - 47.45
8. KS Williamson (NZ) - 47.08
9. DM de Silva (SL) - 46.02
10. HM Amla (SA) - 45.62

Bowling:
1. VD Philander (SA) - 18.12
2. K Rabada (SA) - 18.71
3. DW Steyn (SA) - 19.2
4. PJ Cummins (Aus) - 20.13
5. RA Jadeja (Ind) - 21.94
6. B Kumar (Ind) - 22.02
7. JM Anderson (Eng) - 22.35
8. R Ashwin (Ind) - 23
9. JR Hazlewood (Aus) - 23.21
10. Mohammed Shami (Ind) - 23.56

.. to be fair these still seem to indicate that Williamson and Starc are slightly over-rated.

Standardised averages just since 2016:

Batting:
1. SPD Smith (Aus) - 67.33
2. V Kohli (Ind) - 65.8
3. CA Pujara (Ind) - 52.16
4. Azhar Ali (Pak) - 51.9
5. JE Root (Eng) - 51.73
6. AK Markram (SA) - 50.04
7. Mushfiqur Rahim (Ban) - 48.19
8. KS Williamson (NZ) - 47.64
9. JM Bairstow (Eng) - 47.28
10. KL Rahul (Ind) - 46.79

Bowling:
1. VD Philander (SA) - 16.62
2. JM Anderson (Eng) - 16.65
3. K Rabada (SA) - 18.47
4. PJ Cummins (Aus) - 21.07
5. Mohammed Shami (Ind) - 21.74
6. HMRKB Herath (SL) - 23.06
7. MA Starc (Aus) - 23.36
8. RA Jadeja (Ind) - 23.36
9. BA Stokes (Eng) - 23.4
10. N Wagner (NZ) - 23.98


Starc finally makes an appearance!

 

[Prince EWS]

Wait there's definitely a bug somewhere in that bowling since 2016 list. Numbers are the same as the overall one but with a couple of players removed.

EDIT: Fixed now.

 

[Daemon]

Cribb, did you forget to incorporate the finger spinner factor? Lyon is at 100 right now which is about 800 places too high.

Great work otherwise.

 

[Cabinet96]

Look at Smith's numbers the ****ing monster. Hope he comes back strong.

 

[OverratedSanity]

Matthew Hayden all the way down at 75 was a bit of a shock. Kind of reinforces the view that his statistics flatter him a bit but I was still surprised he ended up that low. Shame Hussain is still lower though.

 

[chasingthedon]

Monumental work Rob. I missed commenting on this originally as I was rather selfishly getting married.

As regards the methodology, I do like that you take into account home and away performance - have you ever thought about a ground scoring difficulty adjustment? I did start something along those lines, however it tends to suffer from a small sample size, though less so now. I did look at incorporating first-class cricket too, and also applied that in my County Points ratings, but tbh the reduction in importance of county cricket has rendered that work moot.

I must say I also like that your batting ratings now show Bradman only significantly ahead, rather than twice as good as he is often portrayed.

I have a problem with that view - if a team has a player who is twice as good as anyone else who ever played, shouldn't they do better than 30 wins and 12 losses?