I think Bubbles addressed all the issues very well, and the solution is a simple and effective one. Basically giving the community more tools to police itself and address issues of stacking as they arise. Whether they choose to do so or not is completely up to them, as it has always been. Not entirely perfect, but for the state of the game I think it's the best choice, especially with Co-Op/Adeventure Mode on the horizon.
And again, I have never been trying to vilify anyone. I know the competitive teams may feel attacked due to claims of team stacking, and some of those people might strike back and say it's the fault of the newer players challenging them. We have to understand that it is a thing that exists and happens, not entirely from the intentions of either party. I'm not aware of any clan that forms a team and starts a lobby with the straight up intention of pub stomping. But as lobbies exist and players form bonds, things will start to tilt in favor of one team over another.
Feel free to stop reading at this point.----------------------------------------------------------------------------------------------------
Even with this issue being resolved, I feel it necessary to post some of the information I gathered. Because frankly it took a lot of time. A lot of time. Too much time. D; (Only managed to get fraction of it done).
First we start off with the design of the survey. The objective was to gather data on matches and look into cases of team stacking and leave rates. This was done by taking a sample every five minutes of currently in progress matches. (You can't have a stacked match if it's not started). This did not include the novice matches (as they cannot be stacked) and was limited to 2v2 matches for simplicity.
Now we have to consider what -is- a stacked match? This definition can vary from person to person. Generally speaking it tends to be a team that sticks together and repeatedly wins by a high margin against opponents that change often (pub stomping). Some might say that it's players of a higher rank, while others would argue that only clanned teams count as stacked. To try and stay as objective as possible, we stuck only with 'rank'. This is not an ideal measure of player skill, but it's a good estimation. Then I set the bar to '40%'. If the difference between the teams weighted average ranks is 40% greater than the avg rank of the lower team, it counts as 'stacked'.
Why 40% and not just some flat number? Because of how leveling and ranks work. Each time you achieve a rank, you have to work much harder to reach the next one. And the more you play, the less difference there is in skill level (So a rank 1 and 2 might be very different in skill, but a 5 and 6 would be pretty close). The average ranks of each team was weighted so that the captain/pilot counted as more of the average. (Put a lvl 13 captain with a lvl 1 crew vs a lvl 1 captain with a lvl 13 crew; assuming they let the lvl 1 actually pilot the ship, the lvl 13 captain is far more likely to win).
Now onto the actual data. Tracking games for 2 hours and 40 minutes, and taking a sample every 5 minutes we come up with this chart:
This is a bar chart showing the weighted ratios of the teams. (HighTeamWeightedAvg/LowTeamWeightedAvg). Anything above 0.4 (40%) is considered stacked. As you can see, this amount fluctuates wildly in both the amount of matches occurring at the same time, and the ratios.
The greatest difference in rank was in this match:
Where you can see that even with the AI dragging down the average (AI's count as rank 0), this team is still much much higher than the opposition. Interestingly, this was not an organized pre-formed team.
Then we have this example of a perfectly balanced team. You'll notice these players cover a wide range of ranks:
And this is my personal favorite, the % of stacked matches occurring at any time:
You add up the number of 'stacked' matches and divide by the total number. Sometimes this goes down to 0, and sometimes all the matches are stacked.
Now obviously we may not all agree to that definition of a stacked team. 40% may be a bit low. Also remember that this is only an isolated example. We'd have to look at the game over a much longer period to get something that better represents the overall state. That's a lot of time for me to sit around and take screen shots of matches every 5 minutes, and with Muse having plans of their own; I won't bother.
Show Posts
