Effects of Strategy on Scored Cargo Distribution

I was re-reading some old threads and came across this post by @SndMndBdy. There is an interesting point here about how the cargo RP threshold should have a direct impact on match strategy. I was wondering if we could identify the effects of the cargo RP threshold on the cargo score distribution to test this hypothesis.

To do this, I collected all 2022 qual match results, I filtered out playoffs because there was no RP in playoff matches. I counted the quintet achieved as a 2 cargo score, to arrive at an adjusted cargo scored for red and for blue in each match. I then took those adjusted cargo scores and made a distribution of occurrences for each cargo count. Here are the raw results:

Adjusted Cargo Scored Distribution
adjusted cargo scored occurrence count
0 524
1 457
2 530
3 620
4 736
5 795
6 812
7 915
8 973
9 987
10 999
11 1004
12 963
13 945
14 919
15 890
16 782
17 769
18 667
19 607
20 1078
21 929
22 745
23 618
24 540
25 475
26 444
27 405
28 343
29 299
30 311
31 242
32 226
33 200
34 174
35 168
36 121
37 103
38 119
39 103
40 86
41 57
42 60
43 55
44 48
45 39
46 33
47 27
48 28
49 16
50 17
51 20
52 19
53 8
54 4
55 5
56 4
57 10
58 2
59 4
60 3
61 3
62 2
63 3
64 1
65 0
66 1
67 0
68 0
69 2
70 0
71 2

Here are those distributions mapped out for the adjusted cargo scored range between 0 and 40:
image

This is really cool to see visually, it is very apparent where the RP threshold is just with a quick glance at this graph. We have some solid proof that teams do indeed target the 20 cargo count. Some of this is likely due to teams pursuing cargo more aggressively at the end of the match when the total scored cargo is below but close to 20. Also the flipside of that, teams after achieving 20 cargo likely played more conservatively, lining up for endgames with extra time to make the hangar RP more likely.

Zooming in to the 10-30 adjusted cargo range, I found a trendline to estimate what the distribution might look like without any RP bonus threshold:
image

Using that line, we can estimate how many additional cargo RPs were achieved this season due to strategy, here are the actual, expected, and difference between actual and expected counts:

Cargo Counts vs Linear Expectation
adjusted cargo scored occurrence count linear expectation actual - expected
10 999 1064 -65
11 1004 1028 -24
12 963 992 -29
13 945 955 -10
14 919 919 0
15 890 883 7
16 782 847 -65
17 769 811 -42
18 667 775 -108
19 607 739 -132
20 1078 702 376
21 929 666 263
22 745 630 115
23 618 594 24
24 540 558 -18
25 475 522 -47
26 444 486 -42
27 405 449 -44
28 343 413 -70
29 299 377 -78
30 311 341 -30

Depending on how you want to count, we can reasonably say that there were somewhere between 200 and 600 excess cargo RPs achieved this season due to strategy effects. Out of the 24096 half-matches in the dataset, this corresponds to an approximately 400/24096=1.7% rate, or roughly 3 extra RPs per event.

As a sidenote, I also found a few dozen matches in which there was a discrepancy between the implied RP award based on the match results, and the RP being actually awarded, those matches are shown here:

RP Discrepancies
event key color Match match Cargo Total cargo Bonus RP Achieved quintet Achieved adjusted total cargo calculated RP achieved
azfl blue 60 15 1 0 15 0
utwv blue 21 0 1 0 0 0
scan blue 15 0 1 0 0 0
azfl blue 41 23 0 0 23 1
mitvc blue 21 20 0 0 20 1
azfl blue 51 20 0 0 20 1
gacol blue 22 19 1 0 19 0
gacol blue 17 19 1 0 19 0
mibel blue 57 16 1 1 18 0
azfl blue 56 18 1 0 18 0
gacol blue 21 17 1 0 17 0
azfl blue 18 17 1 0 17 0
mibel blue 58 17 1 0 17 0
txwac blue 3 0 1 1 2 0
txwac blue 26 0 1 1 2 0
gadal blue 39 0 1 1 2 0
txwac blue 54 0 1 1 2 0
txwac blue 18 0 1 1 2 0
caph blue 15 0 1 1 2 0
scan blue 4 0 1 0 0 0
va305 blue 3 0 1 0 0 0
scan blue 8 0 1 0 0 0
casf blue 21 0 1 0 0 0
inkok blue 5 0 1 0 0 0
tuis3 blue 1 0 1 0 0 0
txwac blue 1 0 1 0 0 0
mxmo blue 33 0 1 0 0 0
txwac blue 67 0 1 0 0 0
txwac blue 10 0 1 0 0 0
mimil blue 31 0 1 0 0 0
txwac blue 31 0 1 0 0 0
azfl red 60 13 1 0 13 0
utwv red 21 0 1 0 0 0
scan red 15 0 1 0 0 0
mike2 red 38 17 1 0 17 0
azfl red 63 17 1 0 17 0
miwal red 30 19 1 0 19 0
azfl red 29 25 0 0 25 1
azfl red 49 19 1 0 19 0
gacol red 37 19 1 0 19 0
ausc red 56 18 1 0 18 0
txwac red 45 0 1 1 2 0
njbri red 10 0 1 0 0 0
txwac red 38 0 1 1 2 0
txwac red 35 0 1 0 0 0
txwac red 50 0 1 1 2 0
txwac red 48 0 1 1 2 0
scan red 5 0 1 1 2 0
scan red 14 0 1 0 0 0
txwac red 39 0 1 1 2 0
txwac red 58 0 1 0 0 0
scan red 9 0 1 0 0 0
txwac red 8 0 1 1 2 0
txwac red 17 0 1 1 2 0
txwac red 37 0 1 0 0 0
mndu red 17 0 1 0 0 0
31 Likes

This distribution is neat and matches a trend in marathons, where runners clearly target a goal time of a round number. These cliffs at those round times are incredibly dramatic.

19 Likes

Asking for @Leap:

How many matches with 17-20 cargo scored by the losing alliance were lost because the alliance pushed to get the cargo RP and sacrificed the climbs?

3 Likes

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.