# Play some balance-testing matches!

Factions uses a skill-adjust algorithm to modify the point value of matches based on the skill gap between the factions. For example, an all-Gold team beating an all-Silver team will be worth fewer points than usual.

So far, we’ve been basing this mostly on guesswork. Why? Well, we just didn’t have the data necessary to calculate how the various tiers perform in a match.

I crunched some numbers using previous arc data and some rather simplistic analysis. You can read the results here.

The fundamental question being asked: what is the relative magnitude of the game-winning impact of various tiers in a mixed-rank game? Let’s say you have 5 Noxian Silvers vs. 5 Ionian Silvers. Then, Noxus swaps two Silvers out for a Gold and a Bronze. Here are three possible conclusions:

• The match is still even, because a Gold is better than a Silver (on average) to the same degree that a Silver is better than a Bronze. (This would include a belief that Golds, Silvers, and Bronzes are really indistinguishable.) Keep the points the same as an all-Silver mirror match.
• Noxus now holds the advantage, because upgrading a Gold to a Silver is a bigger deal than downgrading a Silver to a Bronze. If Noxus wins, it should win fewer points, and if Ionia wins, it should win more points.
• Ionia now holds the advantage, because dropping a Silver to a Bronze is a major change, and upgrading a Silver to a Gold is a lesser change. If Noxus wins, it should win more points; if Ionia wins, it should win fewer points.

Of course, being able to look at two teams and decide which one has the advantage is just the first step: we also need to determine how big of a point adjust should be applied.

## Help Us! FOR SCIENCE.

Here’s how you can help us make Factions more fair. It’s pretty painless!

• Visit the Balance Testing page.
• Join the CCTBalance chat in-client.
• Play some matches! These are custom games with full rosters and regular ban rules: basically, play them like a Ranked or a Normal. Try your best and don’t goof around too much: we’re trying to science here!

This data is delicious in that it is not corrupted by faction strength variables. (Of course, the tradeoff is that potentially ranks work differently in Factions matches.)

## Incentives

It would be glorious if we could get a couple hundred balance matches into the dataset, so we can run the analysis and put together a better skill-adjust formula before the next arc begins.

Here are some incentives:

• For every 3 balance matches that you organize, I’ll give you a bonus vote in the poll to determine next arc’s factions. (One bonus vote has weight comparable to the regular votes of one Summoner.)
• Just playing can earn you prizes, too. For every 50 matches that are sent in, I’ll randomly select one match to bless with my frosting magic. This will include giving everyone in the match 1 bonus vote. There may also be other prizes, as my limited cupcake finances allow. (The problem with being a cupcake is that all the money goes to the bakers, with hardly any for the cupcakes. Curse you, Morgana!) I might also write a story about the match, making up a dispute and giving each Summoner involved a dramatic appearance.
• I pledge that, on any night during Intermission when I manage to play LoL, at least one match will be a balance match. You get to play with me! Isn’t that super-awesome? “Wow, yeah Cupcake, it sure is! You are very good with Warwick and Blitzcrank.” Thanks Naggarok! You’re too kind.

Last, but not least, you will be helping to solve a mystery not only of interest to Factions but also to the general League community.

—CupcakeTrap

Caitlyn, you monster.

Posted in Uncategorized
###### One comment on “Play some balance-testing matches!”
1. StormRevolver says:

im always up for it when im logged in :)