Stat Attack! Batsmen vs the top bowlers - study of last 25 years

[sifter132]

Greetings all! I've been playing around with Statguru and used a feature that I hadn't really tried much before: adding individual players into the engine. By doing this you can find stats pertaining to matches where a certain player was involved. That gave me an idea to look at batsmen who have prospered/failed against teams with good bowlers. Found some interesting stuff, so I thought I'd share!

Since it's my 25 year anniversary of cricket watching, I limited it to batsmen vs the best 25 bowlers of the last 25 years. My cut off was October 1988 - any stats from then on counted. Any time one or more of these 25 bowlers were in the opposition, your runs counted toward the results...if none of these guys were in the opposition then runs didn't count - you were playing 'nobodies'

(The 25 bowlers I used were: Australia - Glenn McGrath, Shane Warne, Jason Gillespie, Mitchell Johnson, Craig McDermott, Merv Hughes, Brett Lee; England - Angus Fraser, Darren Gough, Stuart Broad, Graeme Swann; India - Anil Kumble; Pakistan - Wasim Akram, Waqar Younis, Shoaib Akhtar; South Africa - Shaun Pollock, Dale Steyn, Allan Donald, Makhaya Ntini, Vernon Philander; Sri Lanka - Muttiah Muralitharan, Chaminda Vaas; West Indies - Curtley Ambrose, Courtney Walsh, Ian Bishop.)

Here was my original query so you can play around with it if you wish:
Batting records | Test matches | Cricinfo Statsguru | ESPN Cricinfo


So who won? Well no one 'won', but it was very interesting to see who's average was still good even when playing better bowlers, and who's average crumbled a bit.

Most runs?
Brian Lara. 8961 of Lara's runs came in matches where he was playing against any of the 25 bowlers listed above. Tendulkar 2nd with 8661 runs, Ponting had 7301, Steve Waugh 6713 and Alec Stewart 5th with 6690.

Who played against top 25 bowlers most often?
Michael Atherton. 89.6% of his innings (190 of his 212) came against an attack featuring at least one of those 25 bowlers, poor bugger. Following him were Carl Hooper, Graham Gooch, Alec Stewart and Darryl Cullinan at 80.0%. At the reverse end of the table, Kumar Sangakkara who has played only 77 of his 209 innings (36.8%) against any of the top bowlers. After him comes Samaraweera, then 3 South Africans: de Villiers, Graeme Smith and Amla at 40.5%. General rule of thumb is that 90s batsmen had it tougher than 2000s batsmen.

Best average against the top 25 bowlers?
Damien Martyn - bet you didn't guess that one! Marto averaged 53.29 in 39 matches against these bowlers. Second? Virender Sehwag. bet you didn't guess that one either...53.03. 3rd was Lara 52.4, Michael Clarke 51.18, Ponting 50.35, Graeme Smith 50.11, Steve Waugh 50.09 and Tendulkar 50.06. So only 8 guys average above 50 against the 25 best bowlers, compared to about 20 players who average over 50 against all bowling. Who is the bottom? Tillakartne Dilshan who averaged just 28.3 against the best bowlers (to qualify for this you had to score 10 Test 100s, obviously Chris Martin is lower...) Above him are Marvan Atapattu 32.53, Ashwell Prince 33.7, Ian Bell 34.41 and poor Mike Atherton 35.38

Biggest differences between averages vs top 25 and against nobodies?
Kumar Sangakkara you are a winner sir - step forward. Sanga averages 58.07 overall in Test cricket as of today, but that goes down to 43.54 vs the top 25 bowlers, difference of 14.53. Second? Mohammad Yousuf 14.31 difference. Then comes Jacques Kallis 13.11, Dilshan 12.68 and Samaraweera 12.11. On the other end of the scale, Damien Martyn in the winner: -6.92 difference. Then comes Ramnaresh Sarwan, Sehwag, Jayasuriya and Graeme Smith at -0.55. Mark Waugh and Nasser Hussain are the only 2 others to have averages higher against the top 25, than their overall career figures.

And that leads to the final category: who averges best against non-top 25 bowlers?
Mohammad Yousuf cashes in the best against weaker attacks with an average of 78.37 against the 'nobodies', Kallis 72.22, Sangakkara 67.39, Steve Waugh 66.13, Andy Flower 64.85. No shame in being on this list - if runs are there to be had, then these guys had them. On the other end of the scale, Ramnaresh Sarwan hang your head in shame, only averaging 33.54 against guys outside the top 25. Also poor are Nasser Hussain, Jayasuriya, Damien Martyn and Nathan Astle at 38.69


Conclusions?
First, Damien Martyn - nice job fella. Underrated player.

Second, Ricky Ponting is a guy who usually cops it for not having to play against his own bowlers, therefore having an inflated record. But in reality, Ponting has a good record against other top bowlers. It is Kumar Sangakkara and Jacques Kallis instead who should be copping this claim.

The other guy of note is Virender Sehwag. Labelled a flat track bully a lot, and it may be true. But he's pretty good at smacking good bowlers around, even if it is on flat decks. Contrast between him and his peers Dravid and Laxman is interesting. Sehwag's average goes up against good bowlers, Dravid and Laxman drops. It's the same for other pairs of players: Inzamam played pretty well against the top guys, Mohammad Yousuf was dragged down to earth most of the time; Dilshan averages under 30 against top 25 bowlers, while he predecessor Jayasuriya actually did better against better bowlers; similar effects for Graeme Smith and Jacques Kallis.



Here's the averages for the 61 batsmen who have scored 10 Test centuries in the last 25 years, but if you want to know the stats of any other particular player vs attacks featuring the top 25 bowlers, jot it down and I'll get back to you.

DR Martyn (Aus) 53.29
V Sehwag (ICC/India) 53.03
BC Lara (ICC/WI) 52.4
MJ Clarke (Aus) 51.18
RT Ponting (Aus) 50.35
GC Smith (ICC/SA) 50.11
SR Waugh (Aus) 50.09
SR Tendulkar (India) 50.06
HM Amla (SA) 49.33
MEK Hussey (Aus) 48.68
Inzamam-ul-Haq (ICC/Pak) 48.65
S Chanderpaul (WI) 48.62
Younis Khan (Pak) 48.6
AB de Villiers (SA) 48.03
DPMD Jayawardene (SL) 47.41
GA Gooch (Eng) 46.93
A Flower (Zim) 46.38
R Dravid (ICC/India) 45.63
ML Hayden (Aus) 45.3
AC Gilchrist (Aus) 45.13
PA de Silva (SL) 44.82
Saeed Anwar (Pak) 44.18
RR Sarwan (WI) 43.81
KC Sangakkara (SL) 43.54
KP Pietersen (Eng) 43.46
ST Jayasuriya (SL) 43.23
ME Trescothick (Eng) 43.17
VVS Laxman (India) 43.06
M Azharuddin (India) 42.36
MJ Slater (Aus) 42.28
JH Kallis (ICC/SA) 42.26
ME Waugh (Aus) 42.15
DJ Cullinan (SA) 42.02
RB Richardson (WI) 41.55
AN Cook (Eng) 41.53
MA Taylor (Aus) 41.52
GP Thorpe (Eng) 41.37
JL Langer (Aus) 41.1
SM Katich (Aus) 41.04
MP Vaughan (Eng) 39.97
SC Ganguly (India) 39.86
G Kirsten (SA) 39.64
LRPL Taylor (NZ) 38.97
Ijaz Ahmed (Pak) 38.97
DC Boon (Aus) 38.39
Mohammad Yousuf (Pak) 37.98
PD Collingwood (Eng) 37.53
AJ Strauss (Eng) 37.42
N Hussain (Eng) 37.33
HP Tillakaratne (SL) 37.05
CL Hooper (WI) 36.96
CH Gayle (WI) 36.82
AJ Stewart (Eng) 36.75
TT Samaraweera (SL) 36.65
NJ Astle (NZ) 36.01
HH Gibbs (SA) 35.91
MA Atherton (Eng) 35.38
IR Bell (Eng) 34.41
AG Prince (SA) 33.7
MS Atapattu (SL) 32.53
TM Dilshan (SL) 28.3


Well there's my afternoon done. Enjoy!

 

[asprin]

That's some serious digging there. Sehwag really took me by surprise.

 

[Cricketman]

Great analysis mate! Very interesting stuff. You are right about Sanga. I was almost sad to see his amazing feat of a triple and a century vs Bangladesh in the last test... I could already imagine people on PC 25 years down the line almost holding it against him saying it "was only against a weak bangla attack"

Who has the most hundreds vs the 25 top bowlers? Any home/away analysis?

 

[War]

Hmm i'd select Ryan Harris over Johnson for AUS. Flintoff & Caddick certainly over Broad at least. Fannie De Villiers has a strong case to be chosen over Ntini/Philander. Harsh to not pick at least Hadlee for NZ given the time period you chose.

 

[sifter132]

Great analysis mate! Very interesting stuff. You are right about Sanga. I was almost sad to see his amazing feat of a triple and a century vs Bangladesh in the last test... I could already imagine people on PC 25 years down the line almost holding it against him saying it "was only against a weak bangla attack"

Who has the most hundreds vs the 25 top bowlers? Any home/away analysis?

Oh yeah, should have done most hundreds...Most hundreds is Sachin with 30, Lara 26, Ponting 23, the Waughs next Steve 20, Mark 18. If it's 100s per innings ratio, then Martyn wins with 12 out of 65 innings, Ijaz Ahmed next 8/46, Hussey 10/61, Tendulkar 30/191, Amla 8/51, Sehwag 15/96 then Clarke 13/84

I didn't do any home and away analysis, no. Might do some if I'm bored, but I want to do bowlers next before the Aus-SA series stuffs up the numbers Nor did I look to see whether it was just 1 bowler from that list in the opposition or more than 1 - might have made a difference to the numbers as well. It just started as a curiosity to see how batsmen did when both Warne and McGrath played, and ended up a big spreadsheet

Hmm i'd select Ryan Harris over Johnson for AUS. Flintoff & Caddick certainly over Broad at least. Fannie De Villiers has a strong case to be chosen over Ntini/Philander. Harsh to not pick at least Hadlee for NZ given the time period you chose.

Oh mate, yes the bowling list was VERY debatable In the end it was too hard and I went almost entirely on stats to eliminate my bias! I made a couple of qualifying rules eg. Had to have over 100 wickets since 1988 - that ruled out Hadlee and de Villiers, Malcolm Marshall and Alderman too. The list is basically the top 25 wickets takers of the era, minus the few with the worst averages (Harbhajan, Anderson, Hoggard, Caddick), and replaced with other guys in the 100-200 wicket range who had better averages: Gus Fraser, Ian Bishop, Philander, Shoaib Akhtar.

I definitely think there are better bowlers around eg. I personally don't think Graeme Swann is a top 25 bowler, nor Chaminda Vaas in Tests. Mohammad Asif is a guy I would put in on reflection, bloody talented bowler, even if he was an idiot. Peter Siddle is definitely in the frame as well. Guys with less wickets like Ryan Harris, Shane Bond, Bruce Reid could be argued as well, but I didn't want to lower the wicket number.

 

[War]

Oh ok fair enough @sifter132

 

[Addy]

M.Asif was magician, or an artist just like Glenn McGrath.Alas we are deprived of seeing this guy bowl anymore

 

[blockerdave]

Very good, interesting analysis with interesting results.

Take issue with some of the bowlers, e.g. Leaving out Flintoff certainly, maybe even harmison. Being Anglo-centric, both of those deserve a place in front of Fraser and broad; similarly Ajmal probably deserves a place over Akhtar?

But a very good analysis overall. I'd also wonder how Stewart, Sangakarra and Flower come out of you exclude their matches as keeper?

----------

making those changes (fraser, broad, akhtar out, flintoff, harmison, ajmal, in) you get Imran Khan as top of the averages, with 67.25, but this is from just 18 innings (807 runs) - if you add a 2000 run qualification, Kallis with 13,231 runs at 55.35 is top (as well as top run scorer).

Steve Waugh (10,047 @ 55.20), Tendulkar (12,961 @54.91), Sangakarra (9,531 @ 54.77) and Younis Khan (6,827 @ 52.51) make up the rest of the top 5. Lara is 6th with 10,342 @ 52.53.

Ponting is down at 12th, behind hussey, sehwag, de villiers, dravid and dean jones.

the recently deposed KP is the first English guy, with 7699 runs at 47.52 putting him at 23, with Alistair Cook at 24 with 7,600 @ 47.5

taking runs scored purely as a batsman (not wicketkeeper), and Sangakarra goes to the top of the averages, with 6,414 runs at 66.12

 

[blockerdave]

interestingly, if you do a similar bowling query, who played against 25 "top" batsmen (not necessarily "the" top batsmen, in order to spread nations and eras to get a decent swathe) and a 150 wickets qualification the top 5 by average are

Ambrose, 347 @ 20.25
McGrath, 563 @ 21.64
Donald, 307 @ 21.65
Murali, 772 @ 22.61
Steyn, 352 @ 22.86

with Warne down in 9th behind Pollock, Akram, and Walsh. (Interestingly, Fraser is the first placed Englishman so I was wrong to exclude him above! Prejudiced against that era as I remember all the beatings we got!)

by Strike Rate, it's:

Steyn 352 wickets every 42.0 balls
Akhtar 178, 45.7 (again, wrong to exclude him above!)
Donald 307, 45.9
Younis 271, 47.2
Gough 229, 50.9

Warne is 20th, with 708 wickets every 57.4 balls.

Interestingly (again, apologies for Anglo-centric view) is Andy Caddick who is above Broad by both Average and Strike rate (228 wickets, 29.97 average, 57.9 SR against 238, 30.31, 58.3) so i guess sifter's original query is much better than mine, really needing only Caddick for Broad.

----------

but if you do swap Caddick for Broad, 2000 run qualification, not as keeper, Sangakkara still tops the averages with 5086 @ 64.37

 

[sifter132]

Interestingly (again, apologies for Anglo-centric view) is Andy Caddick who is above Broad by both Average and Strike rate (228 wickets, 29.97 average, 57.9 SR against 238, 30.31, 58.3) so i guess sifter's original query is much better than mine, really needing only Caddick for Broad.

----------

but if you do swap Caddick for Broad, 2000 run qualification, not as keeper, Sangakkara still tops the averages with 5086 @ 64.37

Mate you're all over it - nice tweaking I think that's the beauty of this kind of tool, you can add different players and you'll get varying results. I think originally my problem with Flintoff and Harmison was that it was really hard to take their whole career when they really only bowled well for small parts of their long careers. A batsmen playing Flintoff in 2000 would think it was happy days. I had similar concerns about small samples like Philander, Harris, Stuart Clark, Bond etc. though so that's why I basically left it to stats to make my field.

The interesting thing to me from all this is that in reality the results are quite close ie. it seems playing against the best bowlers narrows the gap between good and bad for the most part, and I hadn't really expected that. I had Martyn at the top with 53 average, but there are stacks of batsmen within 10 of that. It makes you wonder whether the better players aren't actually the players who score consistently when the going is easy Like Kallis and Sangakkara.

 

[StinkyBoHoon]

suprised about sehwags record but it's consistent with why I've rated him so highly.

yes he's a flat track bully, but he's the best flat track bully ever.

----------

I had Martyn at the top with 53 average, but there are stacks of batsmen within 10 of that. It makes you wonder whether the better players aren't actually the players who score consistently when the going is easy Like Kallis and Sangakkara.

this is a good point, I remember there being some thread someone started about how they thought the west indian batting line up of the 80s would do against warne and murali. in my opinion it's really simple, it would depend on the conditions.

on a seaming green top you could play an all time world XI batting line up and providing you picked pretty good swing bowlers they would still be out for a low score most of the time.

averages are indicators of the ability to make significant, large contributions and how large those contributions generally are. they are not indicators of par scores.

cricket is a game in which consistency is measured using some of the least consistent looking numbers in sport. good figures for a series might read 123, 42, 7, 165, 23, 56. that would be a great return in a 3 test series yet contains massively fluctuating numbers. ultimately, if you can bank on your players to score runs when they should that's a massive boon.

one of the reasons I've always been a big fan of jayawardena is he was a huge scorer at home (and as captain) now, sril lanka's away record is definitely poor and jayawardena's record mirrors that. but sri lanka as an emerging nation were in need of a fortress at home first in order to have some credibility and jayawardena provided that regularly scoring daddy hundreds in conditions he knew well. certainly if I was to give bangladesh any type of batsman it would be one that averages around 50 because he averages 70+ at home and in the 30s away. (obviously they need seamers more than anything mind) if bangladesh began winning tests regularly at home then even if their away record had been as it has been over the last 10 years the country would be a much more credible test nation.

 

[used2bcool]

Cricinfo has an article in a similar vein here: Best Batsmen against Top Attacks.

I think the difference here is that the author is not considering just the top bowlers, but the potency of the bowling attack as a whole. He has calculated a weighted average, and assigned a median value using historical data to determine what is a strong attack and what is not.

The only change I would suggest to this analysis is that whether the bowlers were playing home/away should be taken into account. I would propose using home/away averages when calculating the medians that make up the strength of the bowling attacks, since most bowling attacks, as a whole, are better in their home country than abroad.

 

[sifter132]

^Yep saw that article this morning, just coming over to link it Odd that his study makes Sangakkara look the best player of the lot, my numbers are much different for him. Sanga's obviously done well against reasonable attacks but they just haven't featured my top 25 bowlers. I thought it might be because most of the bowlers are from the 90s and Sanga only played top 25 guys 37% of the time, but in that article he has faced 'above average' attacks 38% of the time, remarkably similar...we have just defined them so differently that he averages 56 in that study but only 43.5 in mine!

I'll try and link this page in the article comments and see if the author will elaborate on potential differences, that would be cool

 

[blockerdave]

^Yep saw that article this morning, just coming over to link it Odd that his study makes Sangakkara look the best player of the lot, my numbers are much different for him. Sanga's obviously done well against reasonable attacks but they just haven't featured my top 25 bowlers. I thought it might be because most of the bowlers are from the 90s and Sanga only played top 25 guys 37% of the time, but in that article he has faced 'above average' attacks 38% of the time, remarkably similar...we have just defined them so differently that he averages 56 in that study but only 43.5 in mine!

I'll try and link this page in the article comments and see if the author will elaborate on potential differences, that would be cool

i think, as with my tweaks that also showed sangakkara in a better light, it's fairer to take his record as a batsman separate to a batsman keeper, and that record is outstanding

 

[ethybubs]

The stats include batsman who played on the same team as the bowlers because it says Hayden's top score is 380 but he made that against Zimbabwe. So it is unfairly weighted for people like Hayden, Lara ect. because it is taking stats from almost every game they played, even against terrible bowling line ups.