Tier 10 Battleships: A Statistical Analysis of Current Player Performance

Using a one-way ANOVA, player performance for the tier ten battleships was analyzed based on win rate and average damage as well as the new, unpublished WoWReplays.com Ship Rating and Aggression Rating. Significant differences were found between all ships in all but aggression. While the Yamato is still the top performer, the G.K. is a close second while there is much need for a buff to the Montana.

As many high tier battleship players are aware, there is a noticeable difference between the three current tier ten battleships – Yamato, Montana and Großer Kurfürst (G.K.). As of the writing of this article Warships.Today reports the Yamato as the top performing ship, followed closely by the G.K. then the Montana (Table 1)[1]. Each ship has a slightly difference play style, which has largely been carried down through their respective tech trees. These differences in play style, in the hands of a good captain, should lead to similar performance statistics among the three top tier battle ships if the ships are truly balanced. Most players who have sailed all three, or at least the original two (Montana and Yamato), know that this is not the case and that the Montana tends to perform worse than the other two. This has also been noted by several community contributors in both the EU and NA. But is this a case of confirmation bias? As Table 1. outlines, though the win rates for the top two ships are identical, there is a difference in their damage, but is that difference significant? Further, there is a notable difference between the Montana and the other two ships, like before, are these differences significant? These are the questions I have sought to answer and have analyzed my current data batch to do just that.

The Data

All the data was obtained from WoWreplays.com which is using the Wargaming API to collect data on every player in every ship on both the North American (NA) and European (EU) servers. A statistical program called SPSS was used to download that data from WoWreplays.com and randomly selected 5% of the download data. This equated to over 3,250,000 cases of over 308,000 Players by 223 Ships, meaning for every player ID as many as 223 individual ship stats may be included, in the case of Testers and Community Contributors. From these cases, only those battles that included the ships with ShipID 4179572528 (Großer Kurfürst), 4276041424 (Yamato) and 4277090288 (Montana) [2] were selected for this analysis. This included 10,303 cases with an average of 85 battles in the G.K., 120 battles in the Yamato and 99 battles in the Montana[3]. It is worth further noting that this data is over 34 days old at the time of writing this article, and due to only being a 5% sample, and from both the NA and EU servers, there will be differences than those shown in Table 1.

The Analysis

Using the new filtered dataset of only the three tier ten battleships, as described above, a one-way ANOVA (analysis of variance) was conducted with Post Hoc tests to outline the exact differences between each ship. Both player average Win Rate and average Damage were analyzed along with the untrimmed versions[4] of the new Ship Rating and Aggression Rating (Fig. 1). The results of the ANOVA showed significant differences, beyond a 95% confidence interval, in all areas except Aggression (Fig. 2). Further post hoc analysis (Fig. 3) revealed that indeed, there are significant performance differences between all ships. Our current dataset showed a significant Win Rate difference between the Montana and other two ships, but not between the G.K. and the Yamato. Similarly, the same could be found for the average damage per battle. However, significant differences were found between all ships with the Ship rating, with the Yamato placing first with an average score of 125.96, followed by the G.K. with a score of 119.14 and the Montana being the poorest performing ship with an average score of 108.15.

Conclusion

Without a doubt, there are statistically significant performances differences between all the ships. Though it could be argued that these could be due to the differences in sheer battles played and each either as a result of popularity, ease of play or age in game, the analysis using the current data would nullify any of those arguments. It is now vividly clear that the Yamato still holds supreme as being the most powerful battleship in the game, with the G.K. close on its heels in second place. With time, however, I could see the G.K. contest the Yamato for first, but the current overall meta of the game would have to change for that to occur. Sadly though, the Montana is in a distance third place, easily out matched both the other ships. This can be explained by the fact that both the Yamato and G.K. simply have a better overall damage per minute when you include their secondaries. It is worth noting for the nay sayers however, that even though the Montana is easier to citadel than both the other ships, and has less of a heal; its survivability is not significantly different than the Yamato but significantly better than the G.K (see Fig 4. and Fig 5.).

If we hope to see the Montana become more competitive with the other two tier ten battleships, she will need a significant buff to her performance. This will likely have to be done by boosting her rate of fire. With time, maybe this will happen, but it seems that as far as Wargaming is concerned, the Montana is performing just fine so we will just have to sit and wait…

Table 1.

Battles Win Rate Damage XP
Yamato 791,630 51.23% 88,713 1,667
Großer Kurfürst 292,680 51.23% 85,851 1,576
Montana 579,381 49.17% 75,995 1,573

Figure 1

BB-Fig1

Figure 2

BB-Fig2

Figure 3

BB-Fig3

Figure 4

BB-Fig4

Figure 5

BB-Fig5

[1] https://na.warships.today/vehicles – Only the NA server data, Accessed 5/9/2017

[2] Due to the limitations of SPSS, the WG API’s ship ID had to be used as a defining factor as opposed to the ships actual name.

[3] All means were rounded up to the nearest whole battle.

[4] Other versions of the ratings have been created that ignore players with ships that have less than 10 battles and less than 15 battles. This has been done to limit extremes which have a tendency to skew mean averages, but most importantly the standard deviations.

Author: SnipeySnipes

A technical and analytic writer for the WoWReplays.com blog - I focus on nuances and details that are often ignored or over looked by other World of Warships community contributors. With a background in the social sciences and statistics I aim to answer questions about the game and player population as a whole, backing my findings with numbers as proof. I play World of Warships as often as I can but by no means am a "professional." Though I would say I perform better then the average player, I still screw up and die often! You can find me in game or hanging about the WoWReplays Discord server. Hit me up for questions, comments or suggestions. Snide remarks or toxic comments will be forwarded to my complaints department for disposal.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s