What counts as a large enough sample?

ClutchIQ treats ten or more stored maps as usable and five to nine as a small sample worth flagging. Below five it shows nothing at all. Those thresholds are conventions, not laws — a stable roster and a stable patch matter as much as the raw count.

How do I convert an American price like -110?

For a negative price, implied probability is |price| ÷ (|price| + 100). For -110 that is 110 ÷ 210 = 52.4%. For a positive price, it is 100 ÷ (price + 100); +150 is 100 ÷ 250 = 40.0%.

Does a 65% hit rate mean I win 65% of the time?

It means the line happened in 65% of stored maps in that sample. Future maps are a different sample against different opponents, possibly with a different roster and patch. History is an input, not an outcome.

Guide

Hit rate, implied probability, and sample size

Hit rate is history. Implied probability is the price. Sample size decides whether either is worth reading.

Updated 2026-09-18Hit rateImplied probabilitySample size

Hit rate is history

Hit rate is hits ÷ maps counted. If a player went over 20 kills in 13 of 20 stored maps, the hit rate is 13 ÷ 20 = 65%. Nothing about that number involves a price, a book, or a forecast. It is a count.

Because it is a count, it can be empty. Zero stored maps produces no hit rate at all — not 0%. One or two maps produce a number, but not a trend.

Implied probability is the price

A price encodes a chance. Decimal odds make that visible: implied probability is 1 ÷ decimal odds.

implied = 1 ÷ decimal odds

1.80 → 55.6%. 2.00 → 50.0%. 3.50 → 28.6%.

Decimal	Implied	American	Implied
1.50	66.7%	-200	66.7%
1.80	55.6%	-125	55.6%
2.00	50.0%	+100	50.0%
2.50	40.0%	+150	40.0%
3.50	28.6%	+250	28.6%

A book's price also carries its margin, so the implied numbers across a market add up to slightly more than 100%. That overround is the book's built-in edge, and it is why a fair price is always a little better than the one on the screen.

Subtract them and you have edge

edge = hit rate − implied

65% hit rate against a 55.6% implied price is +9.4 points of edge.

The full explanation, including what edge is not, is in what is edge in esports betting.

Sample size decides whether to believe any of it

A hit rate over too few maps is noise wearing a percentage sign. ClutchIQ uses the following floors so a clean table never implies more confidence than the data supports:

Stored maps	What ClutchIQ shows
0	Nothing. The row is omitted rather than shown as 0%.
1–4	Nothing. Below five maps is an insufficient sample.
5–9	The rate, flagged as a small sample. A rough signal, not a settled rate.
10+	The rate, treated as usable.

Maps and matches are different units

Player props count maps: one row per map played. Series-level outcomes — who won the match, pistol-then-map-win — count matches. Mixed units are hidden by greed, since six maps is not the same sample as six matches.

Questions

What counts as a large enough sample?: ClutchIQ treats ten or more stored maps as usable and five to nine as a small sample worth flagging. Below five it shows nothing at all. Those thresholds are conventions, not laws — a stable roster and a stable patch matter as much as the raw count.
How do I convert an American price like -110?: For a negative price, implied probability is |price| ÷ (|price| + 100). For -110 that is 110 ÷ 210 = 52.4%. For a positive price, it is 100 ÷ (price + 100); +150 is 100 ÷ 250 = 40.0%.
Does a 65% hit rate mean I win 65% of the time?: It means the line happened in 65% of stored maps in that sample. Future maps are a different sample against different opponents, possibly with a different roster and patch. History is an input, not an outcome.