2b6692c1f6
finish scalachess upgrade to 10.0.4
2020-09-21 11:33:48 +02:00
b6c222c919
withFilter tweak
...
replace filter followed by map, flatMap or foreach by withFilter
when filtered sequence isn't option
As a scala compiler do for guards in for loops
2020-09-01 01:27:32 +03:00
1d0e445dbb
more code tweaks
2020-08-16 15:06:40 +02:00
7a2918edcc
cats compile
2020-08-12 08:53:51 +02:00
8e69311178
add move time variance to mod insights
2020-05-28 12:48:20 -06:00
4288bebffd
remove deprecated game blur number
...
replaced long time ago with blur bits
2020-05-27 10:30:35 -06:00
4f9d6637fc
remove unused function
2020-05-06 16:06:03 -06:00
5488a09548
scalafmt 2.5
2020-05-05 22:11:15 -06:00
94cf304465
Merge pull request #6539 from ornicar/unweighted-reportable
...
Use unweighted values on reportable
2020-05-03 09:35:19 -06:00
2bc968f2ec
Remove "reportable && exceptionalDif" condition
...
I didn't know that this happened -- but, this may cause false positives or very borderline cases, and otherwise it's confusing that tweaking `reportable` also affects this. Either way it gives too much importance to `exceptionalDif`, imo, and with the improved `markable` it also should no longer be needed, and just having these as a report is fine.
2020-05-03 17:03:16 +02:00
79e9bbc1b3
Use unweighted values on reportable
...
A bit too many false reports otherwise
2020-05-03 12:39:43 +02:00
aaca0dd063
scalafmt
2020-04-23 00:52:15 +02:00
c14da7a90d
Remove unused import
2020-04-20 16:14:25 +02:00
56c48aaa97
Rename percent* to score* because it's not really % anymore
2020-04-20 16:13:43 +02:00
75d20ed895
Slightly increase move time streak CV
2020-04-20 16:07:21 +02:00
e7036517e6
Slightly reduce markable %
2020-04-20 15:06:19 +02:00
695bc81d69
Increase isGreatUser threshold
2020-04-20 14:58:21 +02:00
6d997354cd
Weighted sum
...
Based on time control (+ for fast, - for slow) and move times (+ for
flat, = for anything else)
2020-04-20 14:40:10 +02:00
54fba1de3d
Only downgrade one assessment point on loss
2020-04-15 11:52:32 +02:00
84c03c8ee2
Remove min/max outliers from sliding
2020-04-15 11:52:19 +02:00
ff1cc70bef
reformat with scalafmt
2019-12-13 08:37:32 -06:00
b7bbfde3c4
fix many more warnings
2019-12-08 10:58:50 -06:00
3d52e6bbf8
migrate round module
2019-12-02 10:41:05 -06:00
4b3d3dc1f3
modules/user compiles
2019-11-29 20:16:11 -06:00
23683e5e70
move hold alerts out of the Player object (untested)
...
Because we manipulate so many player objects,
and because we rarely care about hold alerts,
it makes sense to fetch them through a
dedicated DB query.
2019-08-25 18:03:58 +02:00
93b0775645
Add the CI to sfAvgGiven
2019-08-22 18:15:15 +02:00
e098666de5
suggestions by cchantep
2019-08-22 18:09:14 +02:00
980e99706b
Add a simplistic normal-distribution based CI to mod assessments
2019-08-14 18:20:22 +02:00
3937642922
Remove redundant cheat assessment
2019-07-18 21:13:38 -07:00
b9430b6ff3
increase blur streak requirements
...
it makes too many false positives
2018-12-19 11:41:18 +08:00
df27b4fd90
shorten autoreport texts
2018-12-09 16:53:28 +07:00
74cbbbb9b7
Get rid of moderate move time streaks
...
They don't appear to be useful at all, but they can still color a dot orange in Display.scala.
2018-08-27 11:11:31 +02:00
863cd90f9e
optional assessment mtStreak
2018-08-24 11:55:03 +02:00
29fc322dde
optional assessment blur streak
2018-08-24 11:44:56 +02:00
a4171b862c
Fix StatisticsTest
2018-08-23 11:50:09 +02:00
d046a81ada
Up all values by 0.1s instead of 0.5s
2018-08-22 22:43:41 +02:00
81493097d0
Improve mod view for new streak detection
2018-08-21 19:04:58 +02:00
6349fad72e
Everything in Assessible can be a lazy val
2018-08-21 18:31:25 +02:00
f50d8d264b
Make 'color' a class argument of Assessible
...
All methods were using it anyway
2018-08-21 18:29:02 +02:00
c61b835d5c
disable highly consistent move time/streaks to 1+0
...
too many 1+0 games getting evaluated as :( or >:( because of this. so let's only keep that for games slower than 1+0, where it actually is useful
2018-08-20 12:11:25 +02:00
d2c13e430f
simplify highlyConsistentMoveTimes clock selector
...
which also excludes hyperbullet games from it
2018-08-16 20:24:37 +02:00
6e7bc5755f
use estimateTotalSeconds
2018-08-16 19:45:45 +02:00
a82e5c4579
0+1 games shouldn't count for highly consistent mt
...
The move times there are usually a little flatter than other bullet games, and I've seen a false positive with this. Keeping the streaks though because those are much stricter.
2018-08-16 18:58:18 +02:00
5676ec2a14
move time streaks: 10 --> 15
...
should reduce false positives of this kind
2018-08-15 15:08:58 +02:00
9bf9475b60
code tweak
2018-08-13 03:18:20 +02:00
194b5f316a
try to reduce flat move time streak detection
2018-08-13 03:17:42 +02:00
3617c1eb02
simplify player assessment data
2018-08-11 03:02:06 +02:00
2c28b716b6
evaluation refactoring: move a few methods around
2018-08-01 16:50:07 +02:00
7b41674553
No 'highly consistent move times' for ultra (except for streaks because highly-premoved sequences are excluded anyway)
2018-08-01 16:30:45 +02:00
9fb7484935
Revert "Consistent move time streak assessment: only mid-game"
...
This reverts commit dcb9845f03
2018-08-01 16:11:56 +02:00
Niklas Fiekas