Okay, so this post is going to be long, but I think it’s important enough, because it affects everybody given the new event structure.
What is the actual average score per mode? And how much variation is there between good players and average players in the time to complete score based tasks.
So first we’re going to establish what the average score is, and how it changes with player capability. We’re going to stick to AB/RB for this, as Sim Air is very different in a lot of ways, and Sim Ground has thin data.
I took a randomly chosen full squadron (which shall remain anonymous, and may or may not have been the squadron of a prominent “just git gud, scrub” advocate on these forums, the quality of the squadron or its players has no bearing on this analysis, it’s just a sample data set). I noted down average score, and also the average relative place of all players who had more than 100 games in that mode, for the last month. These are all publicly available stats everyone can see on everyone else’s server record.
I then ran scatterplots on that data to get a sense of how fuzzy it was going to be. My theory was that, because average placement in a match is tied to your average score, there would be, if not a linear relation, a fairly close one. It wasn’t going to be perfectly linear, because in some modes there is going to be wider variation in the game experience between BRs than others.
Here’s the results for the six non-Sim modes. In all cases, the player’s average placement in matches is along the bottom, and their average score is on the left side. I also described the best-fit equation (you can judge for yourselves how good a fit it is by the R-squared value)
Naval modes have less data than the others, but there’s still enough to draw some inference from. Both naval and air have a fairly wide variation in performance, again probably because different BRs have different average score per game generally. Naval ships at higher BRs have more damage to take for instance, which could increase score for higher-rank matches over lower-rank ones. Similarly in Air the relative proportion of score in bases, ground targets, caps, and kills will also vary significantly by BR.
Ground modes actually were surprisingly linear, with very little deviation up or down from the best fit function. This could be consistent with the fact that most ground maps and ground match sizes are roughly the same across BRs.
So what this gives us is a set of data that we can use to evaluate average score for different qualities of player fairly precisely. Picking the values of 50, 67, 75 and 80, corresponding to the percent of players on average who are below you on the scoreboard, we get this table:
Spoiler
| Score/game | ||||||
|---|---|---|---|---|---|---|
| Air AB | Air RB | Ground AB | Ground RB | Naval AB | Naval RB | |
| 50% player | 1073 | 701 | 1156 | 1098 | 1050 | 1002 |
| 67% player | 1644 | 1230 | 1681 | 1621 | 1511 | 1422 |
| 75% player | 2010 | 1603 | 1976 | 1899 | 1880 | 1677 |
| 80% player | 2279 | 1891 | 2176 | 2082 | 2161 | 1859 |
The top-quartile of players (in any given game) is doing about double the average player’s score in all modes. So what happens if we add in standard event multipliers? This:
Spoiler
| Score/game (adjusted) | ||||||
|---|---|---|---|---|---|---|
| Air AB | Air RB | Ground AB | Ground RB | Naval AB | Naval RB | |
| Multipliers | 1 | 1.4 | 0.93 | 1.33 | 1.9 | 2.2 |
| 50% player | 1073 | 981 | 1075 | 1460 | 1994 | 2204 |
| 67% player | 1644 | 1722 | 1564 | 2156 | 2871 | 3129 |
| 75% player | 2010 | 2244 | 1838 | 2525 | 3572 | 3689 |
| 80% player | 2279 | 2648 | 2024 | 2770 | 4105 | 4090 |
To figure out how long it takes players in these events to complete, we need to introduce another stat, max games per hour. This is based on some work I did a while ago, and is across all BRs. It pegged a ground RB to 5 games an hour, and based on server replays, figured out how many games were being played per hour by a similarly dedicated player in other modes. (Unfortunately I didn’t collect data for naval RB, so we’re going to have to drop it from this analysis going forward.)
Max games per hour:
Air AB: 5.4
Air RB: 5.6
Ground AB: 4.9
Ground RB: 5
Naval AB: 5.3
These stats are for the duration of the whole match. Obviously if you don’t play most matches to the end for whatever reason, you could fit in more here, but this is a good set of figures I find for players playing full lineups to their last vehicle, at least most of the time. Basically air and naval games go a bit faster than ground games (note, this probably isn’t the best estimate number for naval, where a lot of really short low-tier games are being mushed in with a lot of really long high-tier games).
Taking that factor into account, here’s the average score per hour, after multipliers per mode, based on player quality.
Spoiler
| Adjusted score/hour | |||||
|---|---|---|---|---|---|
| Air AB | Air RB | Ground AB | Ground RB | Naval AB | |
| Games/hour | 5.4 | 5.6 | 4.9 | 5 | 5.3 |
| 50% player | 5795 | 5493 | 5266 | 7300 | 10569 |
| 67% player | 8879 | 9642 | 7662 | 10782 | 15216 |
| 75% player | 10853 | 12565 | 9006 | 12627 | 18934 |
| 80% player | 12305 | 14827 | 9917 | 13848 | 21756 |
What one can say is AB players are being put at a huuuge, almost prohibitive disadvantage here. A top-fifth player in AB is completing this event 33% slower than top RB player, and even less than a top third RB player by these metrics.
So if we take the current challenge, to get the 22,500 points per day, we’re looking at the following hours per day over the 16 day period (I know, it’s just a ground event, but we’re getting air and naval events after this one where this will be more relevant, assuming a similar event with similar multipliers.)
Spoiler
| Time to complete 22,500 points per day in hours | |||||
|---|---|---|---|---|---|
| Air AB | Air RB | Ground AB | Ground RB | Naval AB | |
| 50% player | 3.9 | 4.1 | 4.3 | 3.1 | 2.1 |
| 67% player | 2.5 | 2.3 | 2.9 | 2.1 | 1.5 |
| 75% player | 2.1 | 1.8 | 2.5 | 1.8 | 1.2 |
| 80% player | 1.8 | 1.5 | 2.3 | 1.6 | 1.0 |
You can all decide for yourselves what a reasonable commitment of hours per day is for a player to obtain a (high tier!) digital tank, and what level of player is being excluded from this event and who it’s “made for.”
Similarly, to have a coupon to trade at the current values, here’s the time per day in hours for that, over the 16 days.
Spoiler
| Time to complete 41,667 points per day in hours | |||||
|---|---|---|---|---|---|
| Air AB | Air RB | Ground AB | Ground RB | Naval AB | |
| 50% player | 7.2 | 7.6 | 7.9 | 5.7 | 3.9 |
| 67% player | 4.7 | 4.3 | 5.4 | 3.9 | 2.7 |
| 75% player | 3.8 | 3.3 | 4.6 | 3.3 | 2.2 |
| 80% player | 3.4 | 2.8 | 4.2 | 3.0 | 1.9 |
Fair? Unfair? Your mileage gonna vary on this one. But those are the actual stats. People claiming they are finishing this in more or less time should be up front about what their average relative position in the mode is, though. Makes a big difference if you’re top-half or top-fifth though, as you can see here.
UPDATE: Okay, some people have brought up what percentage of people we’re talking about in each of the categories above. We can crunch the sample we have (N=208) through a standard deviation calculator, assuming the distribution is normal for these purposes, and see what percent of players are better or worse than a given relative position. The standard deviation on the sample is 12.502, on a mean of 51.827 if you want to follow along at home. What it comes to is:
50% average relative position: Better than 44.2% of all players.
67% average relative position: Better than 88.8% of all players
75% average relative positionr: Better than 96.8% of all players
80% average relative position: Better than 98.8% of all players.
You can then flip this around from a given score and figure out how many players can get it. So for example, a 1500 score in ground RB equates to 61% on average relative position, which works out to better than 76.7% of all players on the same normal distribution curve described above. So approximately one quarter of players are going to have an average score equal or better to 1500 in ground RB.
Thx to Poul for the proofread.
