Best opening guesses in Champ Rival
The first guess of a Champ Rival round is the only one you make blind, and it shows: a good champion leaves you with a handful of options while a bad one barely rules anything out. We have calculated, using the same data and comparison rules as the game, which champions are best to start with. The result: Karma is the best opening guess among the current 173 champions.
How we calculated it
For each champion we simulated what would happen if it were your first guess against each of the other 172 possible secret champions. In every case we applied the game's exact comparison (the seven clues: green, yellow, red and the year arrow) and grouped the secrets that give exactly the same result. With that we measure three things: how many candidates are left on average (the size of the group the secret falls into), how many are left in the worst case and how many different results that guess can produce. The fewer candidates left, the better the guess was. We do not count the chance of getting it right on the first try, which is one in 173.
The ranking: the 10 best opening guesses
- Karma: leaves 2.9 candidates on average, 6 in the worst case, and can produce 89 different results.
- Talon: leaves 3.0 candidates on average, 7 in the worst case, and can produce 85 different results.
- Irelia: leaves 3.2 candidates on average, 7 in the worst case, and can produce 81 different results.
- Syndra: leaves 3.4 candidates on average, 12 in the worst case, and can produce 89 different results.
- Lux: leaves 3.4 candidates on average, 7 in the worst case, and can produce 78 different results.
- Brand: leaves 3.6 candidates on average, 10 in the worst case, and can produce 78 different results.
- Jarvan IV: leaves 3.6 candidates on average, 9 in the worst case, and can produce 81 different results.
- Zilean: leaves 3.6 candidates on average, 8 in the worst case, and can produce 73 different results.
- Vi: leaves 3.7 candidates on average, 8 in the worst case, and can produce 79 different results.
- Lee Sin: leaves 3.7 candidates on average, 8 in the worst case, and can produce 77 different results.
For comparison: an average champion leaves about 7.0 candidates (that is the value of the champion in the middle of the ranking). Starting with Karma cuts the 172 possibilities down to about 2.9.
Why these champions work
A good first guess is a "typical" champion: it shares something with many others in every column, so each clue can come out green, yellow or red depending on the secret and the range of results is huge. Karma, for example, has 2 positions (yellow with either one it shares), was released in 2011 (the champion in the middle by release date is from 2011) and its region, Ionia, has 23 champions, the most of any. That is why it gives 89 different results.
The 6 worst opening guesses are exactly the opposite: Smolder, Bel'Veth, Viego, Briar, Xayah and Tristana. All of them play in a single position (bottom or jungle), are not human and were released in an extreme year (the year arrow points to the same side in at least 76% of cases). With that combination, most of their clues come out red or with the same arrow and rule out very little. The worst one, Smolder, leaves 19.1 candidates on average and up to 40 in the worst case. A champion being famous or striking does not make it a good first guess.
How much information each clue gives
We also measured how much each column tells you on average, in bits (a bit is what you learn from a yes-or-no question whose two answers are equally likely). The higher it is, the more that clue rules out.
- Year (1.40 bits): is exact 11.3% of the time and the rest of the time comes with an arrow that splits the candidates almost in half (44.4% up and 44.4% down).
- Positions (1.26 bits): comes out green 10.4% of the time, yellow 26.3% and red 63.3%.
- Species (1.25 bits): comes out green 16.4% of the time, yellow 17.0% and red 66.6%.
- Gender (1.00 bits): comes out green 51.3% of the time and red 48.7%.
- Range type (1.00 bits): comes out green 49.8% of the time and red 50.2%.
- Resource (0.87 bits): comes out green 70.5% of the time and red 29.5%.
- Region (0.40 bits): comes out green 7.8% of the time and red 92.2%.
Two useful conclusions. First: the year is the most valuable clue, because its arrow always tells you which way to look. Second: the region gives little information on average (it is almost always red), but when it comes out green it is a very powerful clue, because it leaves only the champions from that region (between 7 and 23 depending on the region).
How to continue after your first guess
We also tried a simple computer strategy that knows the whole database: on every turn it picks the guess that cuts the candidates down the most. With it, any of the 173 champions can be solved as the secret in 2.49 guesses on average and never in more than 4: 1 champion on the first try, 88 in two guesses, 83 in three and 1 in four. You cannot work that out in your head, but you do not need to: you have 10 attempts per round.
- Start with one of the top ten in the ranking, not with your favourite champion.
- Look at the year first: its arrow tells you whether you are looking for a newer or an older champion. Pick the next one so that it splits the candidates you have left in two.
- Every green clue fixes a fact; every red one rules out everyone who has it. If few candidates are left, type one that meets all the previous clues: that way you can hit the answer and, if not, you keep ruling out.
- A yellow clue in positions or species means they share only part of it: the secret has some of those values, but not all of them, or has more.
More general tips in Champ Rival strategy and, if you are into the daily challenge, in Daily Champion tips.
Is the best opening guess always the same one?
With the current data, yes: Karma. But the ranking depends on the 173 champions there are now; when new champions are added it can change, and this guide is recalculated from the game data.
Is using this ranking cheating?
No. It only tells you which champion to start with; it does not reveal any secret. It is like knowing that, in tic-tac-toe, the centre is the best square.
Does it also work for the Daily Champion?
Yes, the daily challenge uses the same clues. There is no limit on attempts there, but opening well saves you plenty.