How Run Game counts approaches and success rate
In Run Game, an approach is a log marked either Success or Rejected. Your success rate is successes divided by successes plus rejections, rounded to a whole percentage. If you log 3 successes and 7 rejections, that's 10 approaches and a 30% success rate.
What counts as an approach
Run Game has two outcomes, so there are two things that count:
| Logged as | Counts as an approach? | In the success rate? |
|---|---|---|
| Success | Yes | Yes, in the top and the bottom |
| Rejected | Yes | Yes, in the bottom only |
| A moment where you hesitated and didn't approach | Not logged | No |
A moment you skipped has nothing to log, so it never enters a count or a rate. If you want to track those, write them in a session's notes or your own notes app.
Everything the app labels "approaches" uses this definition: the number in the doorway on Improve, the totals on Stats, a session's count, and the rows in Session history.
The formula, with a worked example
Success rate = successes / (successes + rejections)
Say you go out for three weekends:
| Weekend | Success | Rejected | Approaches | Rate |
|---|---|---|---|---|
| 1 | 1 | 5 | 6 | 17% |
| 2 | 2 | 4 | 6 | 33% |
| 3 | 0 | 3 | 3 | 0% |
| Total | 3 | 12 | 15 | 20% |
Two things show up. The total rate is 3 divided by 15, which is 20%. It isn't the average of 17, 33 and 0 (that would be 17%), because each weekend has a different number of approaches. And a single weekend can say 0% when your overall rate is 20%.
Where you see it in the app
- Stats, right under the room at the top, a Success rate number sits next to Avg / day. With no approaches yet it shows a dash. Below it are your Success and Rejected counts.
- A range control switches the window: Today, This week and All time are shown first, with 3 days and 30 days under More. Stats opens on Today.
- This week vs last shows the approach count and rate for each week.
- 7-day rolling success rate charts the rate over time (the last 30 days, or the length of the range you pick).
- Session history shows each session's approaches and rate, and an open session lists Approaches, Success and Rejected.
- During a session, a line under the count reads like "2 success · 1 rejected · 67%".
The two 1 to 10 ratings you set when logging don't change any of this. Nor does the "What happened next?" choice on a success (got number, texted, date set, went on a date, dating). A success is a success for the rate however far it went. Those fields feed other parts of Stats instead, like the ratings charts and the pipeline.
Why a rate on a small sample is noisy
With a handful of approaches, one outcome moves the number a lot. After 2 approaches, one success is 50%. After 10, one success is 10%.
Statisticians put numbers on that with a confidence interval, a range of rates that fit what you observed. One standard version is the Wilson score interval, which Wikipedia's article on binomial proportion confidence intervals says was developed by E. B. Wilson (1927) and can be used with small samples. A 95% Wilson interval for the same observed 30% looks like this (calculated with that formula):
| What you logged | Observed rate | 95% interval |
|---|---|---|
| 3 successes in 10 approaches | 30% | about 11% to 60% |
| 15 successes in 50 approaches | 30% | about 19% to 44% |
| 30 successes in 100 approaches | 30% | about 22% to 40% |
After 10 approaches, a 30% rate is compatible with anything from about 11% to 60%. After 100, the range narrows a lot, but it's still not a point.
Two limits matter here. This math assumes every approach is the same kind of event with the same chance of success, and real approaches aren't: different people, places, times of day and openers. And the interval only captures random swing, not those differences. So treat it as a sense of how much to trust a number, not a prediction.
How to read your own rate
These are rules of thumb, not findings:
- Count first, rate second. Approaches are the part you control. A rate after 10 approaches says little, and a rate after 50 says more.
- Compare like with like. A rate from a lively bar and a rate from a quiet weekday morning may not mean the same thing, so a lower rate from harder settings isn't automatically worse.
- Watch the trend, not one night. The 7-day rolling chart smooths single bad nights.
- Don't treat the rate as a grade. It's one number built from your own definition of Success and Rejected, and what you do with it is your call. Dating goals you control explains why counts are easier goals than outcomes.
If you want a target number of approaches, how many approaches per week discusses ranges.
What to do next
- Open Stats and find Success rate. Note your count too. Which is bigger, 10 or 50?
- Switch the range to All time, then to This week. Compare the two.
- Set a count goal for the next week before you set a rate goal.
- Log honestly. Choose Success or Rejected for what happened, so the rate stays meaningful.
- Revisit the rate once you have 30 or more approaches, and read it next to your notes.
For the wider routine, see how to track your progress in Run Game and why tracking your approaches helps.
Sources
- Wikipedia, "Binomial proportion confidence interval" (page text read; the Wilson score interval section was used for its formula, its 1927 origin and its statement that it can be used with small samples). The three interval rows above were calculated with that formula at a 95% level (z = 1.96) and are not figures from the article.
- Run Game's own definitions come from the app's code and screens, not from outside research.