So, right now tier listing is based mostly on tournament results. I understand the need to base a list on statistical data, but I think the wrong data is being gathered.
A tier list is supposed to be a list of what characters are viable for tournament play, and how viable they are. A characters viability is a large reason as to why they place where they do in tournaments, this is true, but placing well in tournaments isn't what makes a character viable.
When attempting to study what characters are good, the list should be based on why those characters are good. Until that happens, the tier list will be skewed by character popularity.
The reason a character is good in tournament play is because of how viable they are in specific matchups. Anybody that's been to a tournament will tell you how important picks and counterpicks can be. This is why on highly developed games (like SF2ST) top players end up being able to play a variety of characters (pick counterpick will BREAK you if you don't play it!). MK is good because his specific matchups against the rest of the cast are almost always favorable. Snake is good for the same reason.
That's the "why" that should be behind tier list studies. Current tier list studies look at the end result of the why. Why is it important to observe the actual "why"? Well, because it can reveal things that the current tier list study does not.
It is possible to gather statistics on specific character vs character matchups. Record the match data of the semi-finals and finals of tournaments running under a standard ruleset, specifically what character wins against what character; compile the match data into win/loss ratios, and ding, you have yourself... at least something interesting.
Underplayed characters data will no doubt be skewed at first. Example: It could be a fluke that there's only 20 tournaments recorded with Wario vs MK matchups and that 12 of those 20 matches were won by Wario. Maybe the MK players simply weren't using him to the fullest potential, and still won the tournament. Still, this kind of data will raise an eyebrow or two, which would lead to more Wario vs MK matches, which would make the data more accurate.
After enough time has passed, this kind of statistical study would guide players in a way that current tier list and opinion based matchup threads simply cannot. That's why I think it's worth while.
+Good for metagame.
+More accurate means of determining characters' true potential.
-More data to crunch.
-Low tiered characters data can be skewed or non-existant.
----------------------------------------------------------------------------------
Here's another way to look at it. Think of professional sports. How do you know that a player is good? (The player is the character in this example). You look at things like batting average, number of line-breaks, number of interceptions, ect. Even if the player is playing on a bad team (underplayed character), his statistics will shine through if he's a good player.
Matchup statistics are the closest you'll find to batting averages, flies caught, number of assisted outs, strikes thrown ect.
A tier list is supposed to be a list of what characters are viable for tournament play, and how viable they are. A characters viability is a large reason as to why they place where they do in tournaments, this is true, but placing well in tournaments isn't what makes a character viable.
When attempting to study what characters are good, the list should be based on why those characters are good. Until that happens, the tier list will be skewed by character popularity.
The reason a character is good in tournament play is because of how viable they are in specific matchups. Anybody that's been to a tournament will tell you how important picks and counterpicks can be. This is why on highly developed games (like SF2ST) top players end up being able to play a variety of characters (pick counterpick will BREAK you if you don't play it!). MK is good because his specific matchups against the rest of the cast are almost always favorable. Snake is good for the same reason.
That's the "why" that should be behind tier list studies. Current tier list studies look at the end result of the why. Why is it important to observe the actual "why"? Well, because it can reveal things that the current tier list study does not.
It is possible to gather statistics on specific character vs character matchups. Record the match data of the semi-finals and finals of tournaments running under a standard ruleset, specifically what character wins against what character; compile the match data into win/loss ratios, and ding, you have yourself... at least something interesting.
Underplayed characters data will no doubt be skewed at first. Example: It could be a fluke that there's only 20 tournaments recorded with Wario vs MK matchups and that 12 of those 20 matches were won by Wario. Maybe the MK players simply weren't using him to the fullest potential, and still won the tournament. Still, this kind of data will raise an eyebrow or two, which would lead to more Wario vs MK matches, which would make the data more accurate.
After enough time has passed, this kind of statistical study would guide players in a way that current tier list and opinion based matchup threads simply cannot. That's why I think it's worth while.
+Good for metagame.
+More accurate means of determining characters' true potential.
-More data to crunch.
-Low tiered characters data can be skewed or non-existant.
----------------------------------------------------------------------------------
Here's another way to look at it. Think of professional sports. How do you know that a player is good? (The player is the character in this example). You look at things like batting average, number of line-breaks, number of interceptions, ect. Even if the player is playing on a bad team (underplayed character), his statistics will shine through if he's a good player.
Matchup statistics are the closest you'll find to batting averages, flies caught, number of assisted outs, strikes thrown ect.
