Interesting stuff Ananth.

As you have pointed out, the lower order batsmen are probably at a disadvantage in terms of % of teams runs scored, due to their relative lack of opportunities vs openers.

One potential build on this analysis would be to look at the % of teams runs scored in losing/drawing matches as well and then compare this to the % of teams runs in winning matches. You could then use the difference between them to see which players "overachieved" in winning efforts.
[[
Jeff
The idea is to do specific evaluation of top order batsmen. It is possible to do a special analysis of the late order batsmen who have a BPI index of over 7.0 to see which of these have contributed to wins.
One thing I must say with this blog. I am never going to be short of ideas.
Ananth:
]]

Interesting that the stats here show Sachin to have scored a decent percentage of his runs in victories, yet he is often accused (mainly by those at HoldingWilley) of not having won India enough Tests or having scored most of his runs and centuries in losses. Just shows that ultimately people can be as selective and biased as they want with statistics. It is better to simply look for the top 5 batsmen of the last 20 years across both Tests and ODIs, than attempt to place any batsman higher than the other. On the top 5 it is hard to go past the only 5 batsmen to have scored 10000 runs in both Tests and ODIs. No one would disagree that Tendulkar, Lara, Ponting, Dravid and Kallis are about as fine a collection of batsmen as you can get. Any discussions should simply be left at that.
[[
Avi
I agree with you that Tendulkar's contribution has been significant. However also make a note of Dravid's contributions which are also very good.
Ananth:
]]

AloeVera and Cricket analysis - nice combination

Hello there.
I have something to say, about aloevera, something about which you wrote above, about health and cosmetics... For a long time, I and my friend use the products of the forever living. We every time see the nice results and also we make money for our families and we are happy. My friend works with Aloe Vera in the company of flp has more than 5 years (My friend works ONLY in the FLP Company and has a wife and three children). I know Aloe Vera products for beauty for a long time, but a year ago, began working as a distributor in the forever living products company.
Of course, this job is not easy, but in no other case, you can earn so much much money with so much much fun and smiling faces around..
So if you want to discuss something about which you wrote above, and about my experience with Aloe products for weight loss management, I'll always be glad to talk.
Andrew - Aloe Vera Distributor

There is a big flaw in this analysis. For simplicity, let us consider a 5-match Test series. Now, consider the "Mr. Dependable" of one team, that is, the batsman who is most relied upon to score the runs. Here are his scores and result ...

Game 1: 150 & 75* - WIN
Game 2: 0 & 2 - LOSS
Game 3: 1 & 3 - LOSS
Game 4: 0 & 0 - LOSS
Game 5: 3 & 1 - LOSS

Apparently, according to your statistics, this batsman is really "Mr. Dependable" because he scored more than 95% of his runs in victories, eh?

Arvind,
On the other hand, you are the one who has not understood the analysis.
The ratio is between the batsman runs and team runs in wins and not batsman runs in wins and his total runs as you have understood.
In your example let me say that in the first test the team scored 450 & 300. Then the share of the batsman will be 227/750 which is 30% and very good. It is true that in the wins the said batsman has contributed 30%.
I suggest please read the article carefully one more.
Ananth

yeyyy....... Lara > Dravid > Tendulkar > Ponting. I am so happy now.

Lower order batsmen are certainly at a disadvantage here;they are at disadvantage at any ranking which take runs/averages into account. But whether they scored less cause they batted lower or they batted lower because they scored less is uncertain.

The higher ranking for Sidhu is probably due to the fact that all his victories except for may be one, came at home where once the shine came off, the openers can pile up the runs. And then he was one of the best players of spin.

We must also somehow differentiate between batsmen who got to play 2 innings vs who got 1 innings(highly possible in case of a victory,especially with no 5/6)

Ananth, a request, can you add a column mentioning how many parternships the batsmen was involved in. It is also a team based peer comaprison which I guess will be less biased against lower order batsmen

Love
The problem is that the batsman dismissal information has not been available for test matches other than the last 15 years. So it is difficult to do a partnerships analysis.
Ananth

@lovegoel
difference between lara & other (u mentioned) comes from the total team run in the winning cause..avrg wi team run in lara's 32 won tests is 450 while for tendulkar and dravid, it is 530, for ponting it is 560...so it is very hard for them to match lara's % of team score...had this been lara's credit, he would have won more than only 7 tests (after walsh's retirement) including 2 against zim & ban) in last 6 yrs of his career when he averages was in the higher side. lara's contribution shows actually the relative superior bowling strength of w.i. than its batting strength in his era...u can check that with sobers record...he also scored more runs/won test despite being a middle order batsman but came behind of lara

Ananth-
This is a neat piece of work. Regarding Viv Richards not being higher up in order as also some of his contemporaries, it does not detract from his greatness and is indeeed a tribute to the awesome WI team captained by Clive Lloyd in those years.
Those from India that lived thru the 70s/early 80s probably are not surprised that Vishwanath ranks up there as both he and Gavaskar were the batting mainstays thru that period. Of course Gavaskar's career lasted another 34 matches beyond Vishwanath's. That Gavaskar is at number 5 indicate there were other batsmen who started contributing more in the later 80's until Gavaskar retired ? Perhaps.
Elsewhere, someone has suggested taking a look at the runs percentage in drawn matches particularly for lower order batsmen. I wonder if there could be a measure that separates an inning in a match-saving draw versus a draw on a dull wicket which typically leads to both teams scoring lots of runs.
[[
Youvi
All your points are well made.
If and when I do an analysis based on draws, I will take care to distinguish between draws which were run-feasts with no chance of result and the Cardiff, Napier and Wellington type of draws.
Ananth:
]]

I’ve done a bit more digging on % of teams runs scored in non-winning matches, and the results are very interesting.

Firstly, I should say that I’ve used Statsguru for this and so I can’t exclude extras from team scores as Ananth has done – so my numbers will be slightly different from his.

I wanted to know which batsman had the greatest difference in % of team score in winning matches vs non-winning.

Most players I looked at scored a bigger % of runs in winning matches (obvious for top order players) but there were big differences between players.

Shoaib had the biggest difference: He scored 19.6% of his teams runs (including extras) in wins but only 11.5% in non-wins – a ratio of 1.7

Bradman scored 27.1% in wins but “only” 19.7% in non-wins – a ratio of 1.4.

The only player I looked at who had a ratio under 1 (ie scored a bigger proportion of runs in non-wins) was Lara, who scored 18.6% of his teams runs in wins but 19.1% in other matches
[[
Jeff
As I have mentioned in my reply to Youvi, I will look at draws but do a bit of fine-tuning in terms of selecting meaningful draws as compared to the 1960s type draws scoring at a princely rate of below 2.5 rpo.
Ananth:
]]

Excellent analysis Ananth. I wonder if you can find time to measure effects of toss on end-result in both tests and ODIs so find out which team did best after losing a toss. Do you think Australia would still come up on top? I would love to get analysis done on umpiring errors on end result (where Australia has benefited a lot) but that's impossible so I wouldn't bring that up again. Thanks again.

one word.excellent.i say no more

Where is Viv Richards ? The problem with this analysis is that in a team of very good batsmen a high percentage will be hard to get especially if you have very good openers. If you're in a weaker side like Lara then you account for more runs sicne the rest are mediocre and will most likely fail.
[[
Zack
If a player is in a very good team, he necessarily has to share the credits with other, equally great, players. That does not make their contributions any less nor make this analysis any less.
Let us put it this way. Richards could bat freely with the knowledge of 10 great players supporting him. For the later part of his career, if Lara had 3 players supporting him, it was a surprise.
Ananth:
]]

Individual matches stand out in peoples minds. I can remember Laxmans winning double-century vs Aus, or Greenidges winning double vs Eng or Richards many winning performances vs Eng, Ind.

I honestly cannot remember a single Shoaib match winning performance. All I can remember is the many times he got selected, then dropped from the Test team. Not exactly a ringing endorsement of his match winning capabilities. That attribute should go to his captain, the great Imran Khan who cajoled his team to victory with superior leadership and excellent all-round performances.

Not to say that this analysis lacks merit; some useful tid-bits did emerge. Perhaps a batsman who takes a match and shakes it by the scruff of the neck is different from one who bats well, yet is fortunate to be in a team where the major match-winning contributions may have come from excellent leadership and/or bowling performances from others.

Interesting analysis. Am I right that the L column is identifying left-handers? It does not seem to say this anywhere.
[[
Pelham
Yes. I did not put in a separate legend since it seems obvious.
Ananth:
]]

Engle’s comment made me think whether there is a way to determine the number of times an individual batsman won the match for his team.

If a batsman scored at least 30% of the oppositions combined total, AND he averaged 3 times more than his team mates (batting positions 1 to 7), then you could probably say the individual’s batting contribution won it for his team.

Ananth, if you think there is any merit in this suggestion, is there a way you could determine this?
[[
Deon (and Engle)
As usual you have provided me food for thought. It is a great idea. Using my performance analysis points I can certainly determine a players' contribution to a win. Some fine tuning will be needed.
I will look into it.
Ananth:
]]

I’ve looked at the 12 matches that Pakistan won with Shoaib playing and here are the results and Shoaib’s contribution:

87 vs Eng Inns & 18, Shoaib 16 out of 353
88 vs WI 9 wkts Shoaib 46 out of 435 & 13 out of 32-1
88 vs Aus Inns & 118 Shoaib 94 out of 469
90 vs NZ Inns & 43 Shoaib 203 out of 433
90 vs NZ 9 wkts, Shoaib 105 out of 373 & 42 out of 77-1
90 vs NZ 65 runs, Shoaib 15 out of 102 & 142 out of 357
90 vs WI 8 wkts, Shoaib 86 out of 345 & 32 out of 98-2
92 vs SL 3 wkts, Shoaib 30 out of 221 & 7 out of 188-7
92 vs Eng 10 wkts, Shoaib 55 out of 380 & DNB
93 vs Zim 131 runs, Shoaib, 81 out of 423 & DNB
93 vs Zim 52 runs, Shoaib 18 out of 245 & 13 out of 248
95 vs SL Inns & 40, Shoaib 57 out of 459

So really, Shoaib had one series (1990 vs NZ) where he truly can be said to have played a (if not THE) major role in Pak winning, but in most of the other matches he played at best a supporting role

Great analysis once again!
I was particularly astonished looking at number of wins Ponting has appeared in (90!!), followed by Hayden (71), Kallis (64), Haynes (60) and Tendulkar (51). Lara (32) surprisingly lags way behind.
However, looking at Player-of-the-Match awards, it's Kallis who leads the list with 20, followed by Ponting (15), while Tendulkar and Lara share 12 awards.
And, if you take the ratio of awards/matches won among these 4, surprisingly it's Lara who tops with 37.5%, followed by Kallis (31.25%), Tendulkar (23.53%) and then Ponting (16.67%).
It clearly indicates how much impact Lara has had in those 32 wins! It also shows that Ponting's 90 wins were actually the result of the great teams he was part of (with due respect to his batting and leadership skills).
I'm not sure if it's a good topic for analysis, but would be great to see such numbers for other players as well.
[[
Sushamna
Will look into it. First I must confirm that the MOM award info is available correctly for all ODI matches.
Ananth:
]]

Hi Ananth, I tried to answer my own question about which team does best by losing toss in ODIs because toss is so critical for making most out of ground conditions. I looked at all results since Jan 1990 and you can see only 2 teams have more than 50% success rate after losing toss. In fact, Australia and England do better when they lose toss!!

WMLT WMWT %Won after losing Toss
Australia 177 159 52.68%
England 80 78 50.63%
South Africa 130 134 49.24%
West Indies 93 97 48.95%
Pakistan 142 154 47.97%
Sri Lanka 120 132 47.62%
India 123 155 44.24%
[[
Vijay
Thanks.
Australia I can understand. It reflects their overall quality and the reduced dependency on toss. England, with their overall lack of ODI success, is a surprise. Biggest surprise is India with its dependence on toss.
Ananth:
]]

Ananth - yet another useful article. Winning is more of a team effort and is rather dominated by bowlers in tests. From the table, it appears that the winning side scored, on average, 570 runs in Bradman's case, 546 in Sobers' case, 549 in Sangakkara's case, 475 in Lara's case, 493 in Kallis' case, 490 in Greenidge's case, 410 in Anwar's case, 506 in Chappel's case, 556 in Ponting's case, and 529 in Dravid/Tendulkar's case. This suggests that Anwar's/Lara's contribution either was major in bowler-friendly circumstances or was in matches where their bowlers shut the opponents down cheaply. Again, it would be better to add a column on batsman's average in winning cause (not just runs/match). Also, perhaps, columns on average runs scored when the team's net winning score was 400-450, 450-500, 500-550, 550-600, 600-700, etc.

On MoM point, note that a few of Kallis' awards are mainly as a bowler. A great player!
[[
Alex
Good point.
The Batting average is quite valid in Tests. Let me see whether I can add it to these tables. If not, I promise I will repeat these tables with the averages when I do the bowling tables.
Ananth:
]]

Dear Ananth, this one is also great.

For winning cause sometimes players are unable to prove themselves. Reason is that other players are not upto mark.

Just contrarty to your article who is the biggest looser in batting in test cricket.

Lara is the biggest one with 63 defeats in his career. He broke the record of Alec Stewart with 54 defeats in test.

He scored 5316 runs with the help of 14 centuries and 22 half centuries with batting average 42.19.

Amazing! the man comprises of each and every thing has the worst record although his batting average in defeats is even more than so many batsmen overall batting average in test. It is just like a man with two opposite tastes in his career. He is individaully the best and from team point of view the worst.

He scored 688 runs against Sri Lanka and lost the series by 3 - 0 in which he scored 178, 40, 74, 45, 221 and 130. 226 runs against Australia and lost the match in 2005.

Ananth,
[Unrelated topic - Fielding]
The average age of a team determines to some degree their ability to field.
Could you perhaps publish a list with the oldest sides in Test history (based on average age per player)? The oldest will probably be the England side that toured the WI in 1930, which included two players over 50 and another 4 or 5 over 40.
[[
Deon
Two problems.
One is that I do not have the date of birth information available in my database.
The other is that if you have seen Nehra/Ishant/Panesar/et al field and then see McGrath/Lee/Tendulkar/et al field, age won't necessarily come into the equation.
But I get the drift. Your 1930 English side must qualify for a geriatric specialist to travel with them. Pray consider the 1936 Indian team also.
Ananth:
]]

@ Xolile

The AVERAGE age of that England team in 1930 was 37.5 !!!

You had Gunn & Rhodes who were over 50, Hendren, Haig & Astill who were over 40 plus Sandham & Calthorpe who were very nearly 40.

Bill Voce (20) and Les Ames (24) helped bring down the average age quite considerably. I assume that they must have spent a fair bit of time together socialising on that tour - I can't imagine they had much in common with the rest of the team - most of whom were old enough to be their father.

Of course, this situation was largely driven by the fact that England had 2 seperate teams playing at exactly the same time - the other one was in New Zealand. This was a much younger team (average age 30) with only Frank Woolley over 40.

Dear Ananth another useful article.

Winning is a combine team effort and is either more influenced by bowlers or all rounders in tests like Murli delivers more than 55 overs per match so he has more opportunity.

All-rounders and bowlers have great chances to capture man of the match as compare to specialist batsman. Even if batsman scores 150 or plus in first or second inning of any test then the actual advantage always goes to bowlers or All- Rounders of same team. Normally when batsman scored 150 or plus even if he gets second inning, there is a great chance that he cannot score again big inning in second chance in same test match or even an inning of 50 or more runs.
Second thing is that there is no guarantee that he can get second inning also in winning situation for his team or if gets it but they win the match during his inning and he cannot get his actual score.

This is the main reason why Kallis, Murli and Wasim have more man of the matches in test.

Ananth - re Sushmana's argument ... The MoM awards are not entirely in winning causes. E.g, Lara won MoM for 277 in a drawn match. In decisive test matches, bowlers tend to win the MoM awards whereas batsmen dominate the MoM awards in ODI's. So, MoM should be viewed as a differentiator for batsmen in ODI's only.

Lara played for a weak WI batting line-up. Hence his larger share in the team's runs. In winning causes, he averages 61 while Ponting avarages 62, Kallis averages 63, SRT averages 65, and Dravid averages 66.5. Kallis and SRT have not had to bat in the 2nd innings of these matches whereas Lara almost always not only played but completed his innings in such matches. Hence, the average in won matches is a useful calibrator.

On Viv Richards ... as far as tests are concerned, he sadly became a shadow of his former self 1983-1991. His overall average in matches won is only 53 ... Greenidge and Lloyd average better than him in that respect.

Ananth,
I remember you bemoaning in your previous blog how batsmen hog the plaudits over the bowlers. I think you are falling into a similar trap. The “percentage” of “matchwinning” runs of a batsman relative to other batsmen in the team is often a function of BOTH the quality and form of the other batsmen in the team. We see numerous terrific batsmen with lower percentage of team scores. So, we already know that X,Y,Z were good batsmen…so the only thing this tells us is about the form and quality of the other batsmen in the team.
In ODIs-yes, we can say that it is to a larger extent a batsman’s game, and more often than not it is the batsmen who will dictate terms and actually win matches. But in Tests a batsman, however good he may be, can never “win” a match unless the bowlers get the 20 wickets for a reasonable amount of runs.
so, although both forms of the game depend very heavily on the other team players as far as the final result is concerned…in Tests this happens to a much greater degree than in ODIs
[[
Abhi
I am not quite sure what you are trying to say. If I had compared the batsmen contributions and bowler contributions towards wins in one article what you say makes sense. However these are two independent and mutually exclusive articles as and when I do the bowler share article. I know that 20 wickets have to be taken. So when I do the bowler article the bowlers who contributed to these 20 wickets will get their credit.
Now we are talking about x runs were scored in a match won by the batting team. Who contributed what is the question.
Finally whatever you say, if batsmen do not score enough runs (it might be 100 or 500, it does not matter), the bowlers would not be able to win the matches, as Delhi found yesterday.
Ananth:
]]

Ananth, Jeff,
With Voce bowling and Ames keeping the average age of the "9 men stopping boundaries" was an incredible 41 years. That has to be a record.

Well, let’s put it this way. Let us take the eg. of Lara because he has the highest percentage of modern players. As some other comment I saw above there were very few matches won by WI in Lara’s last few years…and it was in this period that the WI team was the weakest, pertinently in the bowling dept as well.
If you replace Lara by any other modern great: Gavaskar, Richards, Tendulkar, Ponting, Dravid…i.e they feature in the WI team instead of Lara you would get similar stats- Since they would outshine the mediocre batsmen around them and the same team would also be involved in a fair share of wins because of the potent bowling lineup which included Ambrose, Walsh etc.
So, effectively these stats don’t tell us much about the individual batsmen, but more about the rest of the team.
As regards ODIs even with a not so good bowling lineup- who are incapable of too much penetration but are reasonably good at containment backed up with reasonably good fielding…a batsman may well win ODIs practically off his own bat…EVEN if the bowlers do not take all wickets and are even reasonably expensive(not profligate). This of course is impossible in Tests. The bowlers HAVE to actually take 20 wickets and that too at a decent average.
So: the importance of bowlers is relatively LESS towards winning ODIs than winning Tests. Note: Not that they DON’T matter – of course they do, it is relative. i.e a top batsman can actually win ODIs when in a team with a mediocre bowling lineup , but this is virtually IMPOSSIBLE in Tests.
Of course one can give contradictory examples (Delhi) but I’m referring to the general principle involved.

Just to add: No prob with the various analyses. They are great. I was only attempting to “interpret” what we actually glean from this particular analysis.

Ananth,
Further to my earlier comments – to expand a little bit more. Spose a good bowling lineup restricts the opposition to say 220 in an ODI . A relatively mediocre bowling lineup may then restrict the same team to say 280/90. However, given a great innings this total too may be chased down in an ODI.
However, in a Test the poor bowling lineup doesn’t have the good fortune of bowling only 50 overs. And they will be there for a lot longer – giving up more runs with less wickets. So, from the onset the batsman in a poor bowling unit is looking more towards salvaging a draw than a win.
So, perhaps if we look at batsman contributions during DRAWN matches as well, that may give us a greater indication of actual batsman contribution relative to bowler contributions.

I couldn't understand the purpose of the analysis properly. However, from the question "Why ignore a winning 98 or for that matter a winning 48." I assume the analysis is to quantify the contribution of the batsman to the winning cause. If that is the case, there is small problem in "dividing total runs scored in winning" by "total team runs in winning". Here the contributions in the individual matches is getting 'diluted' due to the 'weighted average'. Instead it would be more appropriate to compute individual match contribution percentage and then taking an average of it. This will ensure that batsman whose contribution of 48 to a total score of 102 and still won the match gets due weightage.
(More precisely
if b1, b2, b3 is the individual batsman in winning cause and corresp. team scores are t1, t2 and t3,
your computation is like (b1 + b2 + b3)/(t1+t2+t3)
my suggestion is to (b1/t1+b2/t2+b3/t3)/3.
(percentage is understood)
[[
Unni
I am sure you have understood the idea behind the analysis.
However you have made a very valid point. The point is that there are two ways of looking at things. The ways in which the two of us have approached. And one does not invalidate the other.
(20+40+60)/(160+400+240)=120/800=15%
(20/160)+(40/400)+(60/240)/3=47.5/3=15.8%
There could be another example where the first method leads to a higher result. But my gut feel is that these two methods will be comparable when we consider a number of matches.
I think a better idea is to do both. When I do the bowler article I will try and do both types so that the readers can get an idea.
Once again, a great suggestion.
Many thanks.
Ananth:
]]

Unni,
very good observation. dont think it struck too many ppl. But i think your method is probably more accurate(in principal)..whatever the actual results may work out to be

I was commenting that the 'purpose' is not clear. After reading the article, it was more or less "share of batsmen in their team's winning scores". But I felt the reader's comments and general discussion was on "how much a batsman contributed to team's win". I guess, the second one would be the meaningful purpose for an analysis. If that is the case, I feel both of these methods won't be enough. Generally when people quote the 'count centuries in wins', I guess they really mean that the batsman had topscored and really contributed to win. In that case, there are multiple options to quantify this (without bothering about centuries).
1.Simple one=>how many times the batsman topscored in the wins?(may be divided by total wins)(your "winning 48 runs" will count)
2.More complicated way=>Extend the above by considering number of innnings when the batsman was second, third etc(upto 11th)and compute a 'weighted win position factor'(So, if it is 1.00 he always topscored in all the team's wins)
[[
Unni
This is the second time you have made a very valuable suggestion. Thanks. Will look into it.
Ananth:
]]

Unni,
as ananth says again your observation is unique , as in only you have obtained a new insight. what you say is something to the effect of what i feel (though not entirely). but you have worded it much better.