2 replies to this topic

[Valcoer]

I have to admit that I am not entirely sure how elo works I couldnt follow the math. But it seems to me ( and I may be entirely wrong) that the weight class of the mech and its firepower should be contributing factors for elo evaluation. When elo is used in chess both players have exactly the same pieces with exactly the same capabilities and the only true difference is who goes first and who is better at chess. Between two players that are of equivalent skill that first move can mean a win or a loss.
Applying elo across weight classes and with out taking into consideration the firepower of the mechs is like ranking two equivalent chess players while one is playing at odds of a knight.

The match score also seems to be little more than a way of saying this player did this much damage and that player did this much with little else being used in the determination of how the players performed in a team game. yes sometimes a player with less damage will have a higher match score but the damage done in those cases are ussually less than a hundred points of damage when both have over 400 points of damage done. So I feel that mech weight and fire power should have a huge impact on the final match score. The combined firepower of the team should be reflected for instance a team of splat cats have an average firepower of 90 while a team of standard K2 have an average firepower of 35 (give or take a point) the team of splat cats should have reduced matchscores per point of damage done while the K2 team should recieve substantially higher match scores for each point of damage done. This reflects the skill involved in piloting mechs. In this case two teams of equal skill the splat cats will almost always win as the amount of damage per second is substantially larger. So where is the reward for trying to win with inferior mechs that require more skill to pilot? how can you say that the team that chose the splat cats has any greater skill than the pilots that choose to handicap themselves by using mechs that are far less powerful just because they won the fight? winning is important for planet capture yes, but elo must be about the players not the mechs that they choose to pilot.

In conclusion I would ask this review the data look at the types of mechs that are winning look at the 8 man configurations that you are seeing and ask yourselves do you want the elo to be based on the mechs being piloted or do you want it to be based on the skill of the pilots? if your answer is on the mechs then everything is fine. If you want it to be on the pilots then you must take into consideration the fire power and weight of the mechs in the matches in order to acurately gage (apparently a synonym for gage is filtered) the pilots.

Edited by Valcoer, 05 March 2013 - 05:51 AM.

[Peiper]

Besides firepower, you have to take into account tonnage (an Awesome is NOT equal to an Atlas, even with the same firepower), and equipment (ECM and Streaks give players that have them an 'easy' button, which probably throws off accurate recording of a player's actual skill level). Furthermore, play style is probably ignored. Some players are better team players than others. For example, some players, once disarming a mech, will go on to hunt other, dangerous mechs who are still shooting their teammates where others, more concerned with the K-D Ratio will continually batter away at the disarmed mech while their buddy is being mercilessly pounded by a still armed opponent. Making ELO more transparent MIGHT help me become a better player - the same way showing how much friendly damage I dished out compared to enemy damage might expose to me that I need to be more careful, or that I"m too timid.

I can see how the devs would also hide their matrices to prevent people from abusing the same system. What if I could find a way to pad my own scores to continually be set up with newbs so I can keep kicking their butts? That would be seriously counter-productive.

And, Valcoer,"gauge" is the word you're looking for.

[Valcoer]

those are other instances I had not considered and I have to agree with you that they need to be taken into acount as well. thank you for adding those.