Oliver Stevenson


Hello, thanks for visiting!

I was recently a PhD candidate in the Department of Statistics at the University of Auckland, where I spent my time investigating new statistical methods and models in the field of sports analytics. My research focussed on the development and implementation of statistical models that can be applied to sport of cricket, with the primary aim of estimating and forecasting the past, present, and future abilities of professional cricketers, and to predict the likely outcome of upcoming matches. These models have been proven to provide more accurate predictions of future player performance than traditional cricketing statistics, such as batting and bowling averages.

At present, I work as a Data Scientist at Luma Analytics, a data and analytics consultancy based in Auckland, New Zealand. I also continue to maintain the cricket-related statistical models I developed during my PhD and provide independent statistical and data consulting services to those looking for help with data modelling, data analysis and other data-related queries. To date I have advised clients who work in a range of fields including financial services, information technology, ERP solutions, market research, clinical trials and sports analytics, and I’m always up for a new challenge.

Feel free to take a look at my past and present research below, or get in touch if you have any questions regarding statistics, data, cricket or otherwise.


Research interests:

  • Sports analytics
  • Bayesian inference
  • Machine learning
  • Computational statistics

Doctor of Philosophy (2017 - 2020)

Between 2017 and 2020 I completed a Doctor of Philosophy at the University of Auckland under the supervision of Dr Brendon Brewer. Following on from my Masters, my research involved developing a range of statistical models that can be applied to the analysis of sporting data, with a particular focus on cricket. As with any sport or profession, we shouldn’t expect a player to perform with some constant ability throughout their entire career. Instead, we are likely to observe both short and long-term variations and fluctuations in ability due to the likes of age, experience, fitness and general improvements or deteriorations in technique.

As part of my PhD research, I derived a range of statistical models that employ machine learning algorithms to better estimate and predict the past, present, and future batting and bowling abilities of professional cricketers. When estimating current player ability, these models account for a range of external factors, including recent form, strength of opposition, and venue of past performances (e.g. home or away). These estimates have been proven to provide more accurate predictions of player peformance, compared with traditional cricketing metrics, such as batting and bowling averages. Additionally, the models have the benefit of maintaining an intuitive cricketing interpretation, unlike other popular ranking metrics, such as the official ICC rankings. Given two proposed playing XIs, it is then possible to feed the estimated player abilities into a custom built match-simulation engine to provide a probabilistic prediction of either team winning a match.

Click here to read thesis titled “Form is temporary, class is permanent: Statistical methods for predicting the career trajectories and contributions of players in the sport of cricket”.

See below for an overview of my current Test batting and bowling rankings, including an interactive application that allows users to visualise the batting and bowling career trajectories of international cricketers.

Master of Science (2016 - 2017)

In 2017 I completed my Masters degree under the supervision of Dr Brendon Brewer. My research looked to tell a more meaningful story behind a cricket player’s batting average. Using Bayesian statistical techniques, I explored more in-depth methods of quantifying a cricketer’s batting ability than the simple batting average. More specifically, I built statistical models which estimate how well a batsman is playing at any given point in their innings, allowing us to quantify the cricketing concept of a batsman ‘getting their eye in’. The primary focus was on Test match cricket, with wider applications to 4-day first-class cricket. Using these models, I also explored the plausibility of popular cricketing superstitions from a statistical point of view, such as the commentator’s favourite, the ‘nervous 90s’.

ABSTRACT: Cricketing knowledge tells us batting is more difficult early in a player’s innings, but gets easier as a player becomes familiar with the local conditions. Using Bayesian inference and nested sampling techniques, a model is developed to predict the Test match batting abilities of international cricketers. The model allows for the quantification of players’ initial and equilibrium batting abilities, and the rate of transition between the two. Implementing the model using a hierarchical structure provides more general inference concerning a selected group of international opening batsmen from New Zealand. More complex models are then developed, which are used to identify the presence of any score-based variation in batting ability among a group of modern-day, world-class batsmen. Additionally, the models are used to explore the plausibility of popular cricketing superstitions, such as the ‘nervous 90s’. Evidence is found to support the existence of score-based variation in batting ability, however there is little support to confirm a widespread presence of the ‘nervous 90s’ affecting player batting ability. Practical implications of the findings are discussed in the context of specific match scenarios.

Click here to read thesis titled “The nervous 90s: a Bayesian analysis of batting in Test cricket”.

Bachelor of Science (Honours) (2015)

ABSTRACT: At a glance, data is more meaningful when presented in graphical form. This project explored innovative methods of automating the display of catch data for large-scale conservation projects. High priority was given to developing methods that allow users to interact with their data, affording them some control over the graphics that are produced. Two interactive applications were developed that allow conservation volunteers to select the data they want to view and how to view it. After a day in the field, volunteers are able to use these applications to see their day’s work summarised on a map or graphic. These graphics highlight the positive impact their efforts are having on the local environment, keeping volunteers motivated and engaged in their work. Various methods of improving the automation of these graphics are outlined, as well as other practical uses of these statistical applications.

Click here to read dissertation titled “Graphical applications for large-scale conservation projects”.

Top 100 Test batsmen


Players must have participated in a Test match since 2019 and must have batted in a minimum of 15 Test innings to be ranked.

Last updated 1st October 2021.

RankPlayerCountryInningsRunsCareer averagePredicted averageICC rating (#)
1KS WilliamsonNZ148723054.055.1901 (2)
2SPD SmithAUS139754061.854.8891 (3)
3M LabuschagneAUS31188560.850.7878 (4)
4JE RootENG198922150.150.4903 (1)
5Fawad AlamPAK2289547.150.1640 (20)
6SC WilliamsZIM27103441.448.9603 (28)
7AD MathewsSL161623644.545.0618 (25)
8RG SharmaIND72290946.244.6813 (5)
9LRPL TaylorNZ183747846.244.5693 (13)
10V KohliIND160767151.143.6783 (6)
11DA WarnerAUS159731148.143.5724 (8)
12HM NichollsNZ58226442.742.7714 (10)
13Mohammad RizwanPAK2691441.542.4651 (18)
14FDM KarunaratneSL139517638.642.3712 (12)
15Babar AzamPAK63236242.241.9749 (7)
16CA PujaraIND150642945.641.6673 (15)
17Abid AliPAK2391743.741.1636 (21)
18Tamim IqbalBAN119467339.940.9605 (27)
19Azhar AliPAK166664142.640.7636 (21)
20Mominul HaqueBAN83334942.940.0596 (30)
21TWM LathamNZ103405641.438.9677 (14)
22Asad ShafiqPAK128466038.238.3561 (38)
23RR PantIND40149040.338.3690 (13)
24LD ChandimalSL113415840.038.2526 (49)
25UT KhawajaAUS77288740.738.2NA
26BA StokesENG130463137.037.2664 (16)
27TA BlundellNZ1757238.137.2519 (53)
28HE van der DussenSA1761038.137.2542 (46)
29DM de SilvaSL64225336.937.0588 (32)
30MA AgarwalIND23105245.736.9571 (34)
31F du PlessisSA118416340.036.8570 (35)
32D ElgarSA120434739.236.6661 (17)
33Shakib Al HasanBAN103373538.936.5553 (40)
34Mushfiqur RahimBAN132449036.536.4626 (23)
35BOP FernandoSL2067739.836.3499 (59)
36BRM TaylorZIM58206038.135.5603 (28)
37Sikandar RazaZIM33118736.035.2523 (51)
38AK MarkramSA47182439.735.2651 (18)
39TM HeadAUS31115339.835.2618 (25)
40MahmudullahBAN94291433.135.0540 (45)
41Shadman IslamBAN1547734.134.9NA
42Nazmul Hossain ShantoBAN1754934.334.7338 (98)
43Q de KockSA89324539.134.5717 (9)
44CR ErvineZIM36120835.534.3522 (52)
45KL RahulIND66225835.334.0559 (39)
46N DickwellaSL80244332.634.0589 (31)
47JA BurnsAUS40144237.033.8497 (61)
48AM RahaneIND130474240.233.6630 (22)
49PSP HandscombAUS2993438.933.6NA
50ARS SilvaSL2370235.132.7NA
51JC ButtlerENG92280033.332.6564 (37)
52JO HolderWI91243432.032.6567 (36)
53BKG MendisSL91302234.732.6514 (44)
54RA JadejaIND80211834.732.6529 (48)
55C de GrandhommeNZ39120735.532.5549 (41)
56TD PaineAUS57153432.632.3555 (39)
57T BavumaSA73209732.331.7518 (54)
58HDRL ThirimanneSL81206327.531.5498 (60)
59KC BrathwaiteWI138428332.731.4543 (43)
60Sarfraz AhmedPAK86265736.431.3NA
61Faheem AshrafPAK1859434.931.1517 (55)
62J BlackwoodWI75216830.531.0576 (33)
63Liton DasBAN42122929.330.7509 (57)
64RJ BurnsENG51165732.530.6619 (24)
65SE MarshAUS68226534.329.7NA
66MS WadeAUS63161329.929.6501 (58)
67MDKJ PereraSL41117730.229.4485 (65)
68GH VihariIND2162432.829.3480 (67)
69JL DenlyENG2882729.529.3461 (70)
70RW ChakabvaZIM43106127.229.0427 (77)
71JL PattinsonAUS2541726.128.5NA
72DJ MalanENG2779429.428.4376 (91)
73JM BairstowENG136434433.928.1461 (70)
74Shubman GillIND1541431.828.0459 (72)
75DM BravoWI102353836.128.0380 (88)
76SO DowrichWI61157029.127.7476 (66)
77OJ PopeENG3288231.527.6528 (49)
78SO HetmyerWI3083827.927.4NA
79Shan MasoodPAK47137829.327.3495 (62)
80SSJ BrooksWI1542228.126.7424 (78)
81DP SibleyENG39104228.926.5473 (67)
82BT FoakesENG1641031.526.4336 (99)
83CR WoakesENG60132127.526.3382 (87)
84Haris SohailPAK2784732.626.2396 (84)
85DW LawrenceENG1535427.226.1369 (93)
86R AshwinIND111268527.725.9399 (82)
87CT BancroftAUS1844626.225.9NA
88RL ChaseWI75199628.125.4434 (74)
89Soumya SarkarBAN3083127.724.9378 (89)
90WP SahaIND52125129.124.8334 (101)
91KR MayersWI1655336.924.6428 (76)
93MJ SantnerNZ3276624.724.1355 (97)
94MR MarshAUS55126025.224.1NA
95KK JenningsENG3278125.224.1NA
96PS MasvaureZIM1534624.724.0NA
97JA RavalNZ39114330.124.0NA
98SM CurranENG3781524.723.8391 (86)
99KOA PowellWI83211325.823.5360 (96)
100JD CampbellWI3064023.723.5393 (85)
Top 50 Test bowlers


Players must have participated in a Test match since 2019 and must have taken a minimum of 30 Test wickets to be ranked.

Last updated 1st December 2020 (updated rankings coming June 2021).

RankPlayerCountryWicketsCareer averageAdjusted career averagePredicted adjusted averageICC rating (#)
1JJ BumrahIND6820.323.024.3779 (9)
2PJ CumminsAUS14321.824.325.1904 (1)
3I SharmaIND29132.432.725.7729 (17)
4N WagnerNZ20626.627.625.8843 (3)
5K RabadaSA19722.925.026.8802 (6)
6JM AndersonENG53825.426.527.2781 (8)
7TG SoutheeNZ28429.030.327.6821 (4)
8JR HazlewoodAUS19526.228.128.0769 (11)
9JL PattinsonAUS8126.226.928.4333 (43)
10SCJ BroadENG51327.529.928.5845 (2)
11Mohammed ShamiIND18027.429.428.6749 (13)
12R AshwinIND36525.427.729.5756 (12)
13Mohammad AbbasPAK8021.725.329.6749 (13)
14TA BoultNZ26727.628.729.7770 (10)
15MA StarcAUS24427.029.129.8797 (7)
16BA StokesENG15831.431.630.1587 (24)
17C de GrandhommeNZ4731.630.731.1476 (32)
18Shakib Al HasanBAN19331.130.731.2NA
19KM JarvisZIM4629.433.831.4347 (41)
20KAJ RoachWI19628.229.131.8744 (15)
21Yasir ShahPAK22430.730.932.2609 (23)
22RA JadejaIND21324.627.632.3722 (18)
23UT YadavIND14430.532.732.4650 (21)
24CR WoakesENG11229.330.032.9666 (19)
25MJ LeachENG3429.029.633.0372 (40)
26MA WoodENG5032.434.533.8406 (37)
27NM LyonAUS39031.632.733.8742 (16)
28ST GabrielWI13931.232.434.0663 (20)
29Hasan AliPAK3128.931.234.3NA (NA)
30Taijul IslamBAN11433.233.834.6585 (25)
31Shaheen Shah AfridiPAK3531.329.434.7459 (33)
32KA MaharajSA11033.233.935.5548 (27)
33Mehedi Hasan MirazBAN8733.134.035.8570 (26)
34JO HolderWI11127.527.636.0810 (5)
35MR MarshAUS4238.636.436.1310 (51)
36RL ChaseWI6941.138.336.8505 (31)
37AU RashidENG6039.835.037.9NA
38MM AliENG18136.637.238.5516 (28)
39RAS LakmalSL15137.436.838.5620 (22)
40SM CurranENG4132.135.039.3403 (38)
41MDK PereraSL15635.334.740.2512 (30)
42MahmudullahBAN4246.040.640.8217 (65)
43L EmbuldeniyaSL3039.839.142.6318 (49)
44CBRLS KumaraSL6736.637.843.0514 (29)
45JC ArcherENG3831.136.544.2383 (39)
46MJ SantnerNZ3944.740.944.3262 (58)
47MJ HenryNZ3050.246.046.0249 (59)
48Sikandar RazaZIM3241.841.846.1323 (46)
49AD MathewsSL3352.945.447.4130 (81)
50A DananjayaSL3324.830.347.6NA

*incomplete data is available for players who made their Test debut prior to 2008.

The batting and bowling rankings are obtained from the models developed as part of my Masters and PhD degrees at the University of Auckland. More detail about these models can be found in the papers I have published, which can be found in the Publications section below.

Batting rankings

When estimating a player’s current batting ability the model accounts for recent form, venue of matches played in (i.e. home, away or neutral) and whether the player was batting in their team’s first or second innings of the match. Additionally, the model accounts for the “getting your eye in” process for each individual player. The data support the general belief that players tend to score more runs when batting in their team’s first innings of a match, at a home venue.

The “predicted average” is the the number of runs we expect the player to score in their next Test innings, assuming their next innings is played at a neutral venue and it is unknown whether they are batting in their team’s first or second innings of the match. The official International Cricket Council (ICC) ratings (and world ranking #) are also provided for comparison. The ranking of players is generally similar between the two methods, although there are a couple of notable differences.

Firstly, our model rewards players who are able to overcome the “getting your eye in” process and remain on a not out score, while the ICC ratings simply provide not out innings with a “bonus” that we susepct is too low. For example, Rohit Sharma has a number of 50+ not out scores, suggesting he frequently overcomes the difficult “getting your eye in” process, but for various reasons, has not had the opportunities to convert these not out innings into big scores.

Secondly, the ICC ratings tend to place more emphasis on recent innings compared with our models. Our general findings suggest that there is little evidence to suggest that recent form is a significant predictor of current batting ability for the majority of players. Instead, we believe a player’s underlying ability tends to change slowly over time, rather than erratically between innings as a direct result of recent performances. It is unclear whether the ICC ratings attempt to provide predictive accuracy of ability, or instead tries to formalise expert judgement about who is in and out of form. These two goals may not be entirely compatible.

Thirdly, the ICC apply a rating decay system, whereby players who have not batted in a recent match, see their rating slowly
decline. While well intentioned to reflect the best current batsmen in the world, the decay system is inherently biased as it tends to affect smaller nations disproportionately, who are generally afforded fewer opportunities to play Test cricket.

Finally, while both methods provide a general indication of batting ability, by measuring underlying batting ability in units of a batting average, rather than arbitrary “rating points”, we are able to maintain an intuitive cricketing interpretation when comparing players. Instead of concluding “Steve Smith is 99 rating points better than Kane Williamson”, we can make more meaningful probabilistic statements, such as “we expect Steve Smith to outscore Kane Williamson by 10 runs in their next respective innings”, or “we expect Steve Smith has a 59.9% chance of outscoring Kane Williamson in their next respective innings”. In both statements we are assuming a neutral venue and it is unknown whether they are batting in their team’s first or second innings of the match. Of course, we can update these estimates to include match-specific information, if we know the venue of the next match and whether the player is batting in their team’s first or second innings of the match.

Bowling rankings

Like the batting rankings, the model used to predict current player bowling ability accounts for recent form, venue of matches played in (i.e. home, away or neutral) and whether the player was bowling in their team’s first or second innings of the match.The bowling model also adjusts for the historic strength of opposition batsmen they have bowled to.

In order to make this adjustment, a new metric, “adjusted bowling average”, has been derived, representing a player’s career bowling average if they had bowled to the “average” Test match batsman their entire career. Introducing this metric allows for a fairer comparison of bowling performances, as the relative strengths of batsmen bowled to is accounted for. The derivation of this metric can be found in my PhD thesis in the Publications section below (as of 18th December 2020).

The “predicted adjusted average” is the the number of runs we expect the player to concede before taking their next wicket, assuming they are bowling to the “average” Test match batsmen and the match is being played at a neutral venue. The official International Cricket Council (ICC) ratings (and world ranking #) are also provided for comparison.

Visualising player ability

Here you can find an application that allows users to visualise how these model estimate changes in player batting and bowling ability, on two scales:

1. Long-term changes in batting and bowling ability that occur between innings, over a playing career, providing an estimate of a player’s career trajectory to date, as well as a prediction for their current and future ability. These estimates are what are used to compute our batting and bowling rankings rankings.

2. Short-term changes in batting ability that occur during an innings due to the “getting your eye in” process.


Stevenson, O. G., & Brewer, B. J. (2021). Finding your feet: A Gaussian process model for estimating the abilities of batsmen in Test cricket. Journal of the Royal Statistical Society: Series C (Applied Statistics), 70(2), 481-506. Online version.

Stevenson, O. G., & Brewer, B. J. (2018). Modelling career trajectories of cricket players using Gaussian processes. In R. Argiento, D. Durante, & S. Wade (Eds.), Bayesian Statistics and New Generations: Proceedings of the 2018 Bayesian Young Statisticians Meeting (pp. 165-173). Springer, Cham. Preprint.

Stevenson, O. G., & Brewer, B. J. (2017). Bayesian survival analysis of batsmen in Test cricket. Journal of Quantitative Analysis in Sports13(1), 25-36. Preprint.

Stevenson, O. G. (2017). The Nervous 90s: A Bayesian Analysis of Batting in Test Cricket. Masters thesis, University of Auckland. Online version.

Blog & News

RNZ interview with Nathan Rarere

I recently had the pleasure to chat with RNZ's Nathan Rarere about the Blackcaps' chances of making the World Test Championship Final earlier this week. At present, my models give New Zealand an 83% chance of going to the big dance in June later this year --- go the...

read more