User blog:Homuhomu123/Evaluation of Radar Modifiers on DA Chance

Intro to the Exp't
This is my 6th part of the ASTRE project, in which 1 exp't was done in 2 days. The exp't was designed as comparison with Exp't F, using it as reference, and trying to estimate different radar modifiers on DA chance. Since DA activation chances are generally lower than CI, I used a larger sample size (~300 Attack Chances). According to limited time & resources, I used fatigue leveling method at 3-2a. Besides, test ships will tank damage, but HP always beyond 50% (moderate damage).

Information About the Exp't
- Fleet & Test ship condition:

- Enemy condition: - Test ship & equipment:
 * Morale <20 (severe fatigue / red face)
 * Always beyond moderate damage (中破)
 * Line ahead
 * AS+ achieved for every battle
 * No Saiun equipped
 * Either Line Ahead or DL.



[ Exp't K ]
- Completed in 2 days (2h + 2h + 1h + 3h)

- # Battles & Engagement Form counting

- Fleet total LoS:  448 -> 458  (+12% CI Chance)

- [RM%] = Radar Modifier (Equals 0% in estimate calculation). - #Buckets: 51

- #Non-S:  40

Comparative Study on the Effects of Different Radars (DA)
Using test results from Exp't F. Test ships were kept w/ the same placement, using same equipment (except radar), very close FLoS, and having negligible level differences.

Brief Conclusion:
- Comparative:
 * Radars, especially those with good stats, will marginally increase a ship's DA chance.


 * Between the 2 surface radars,
 * 1) 32 Large  (+8 Acc, +10 LoS) & 
 * 2) 33 Small (+5 Acc, +7 LoS), #32 gives up to 8% higher DA chance than #33.
 * Assuming radars do not reduce DA chance, #33 boosts DA chance by 0% ~ 6%.
 * Assuming data were not off the 90% interval, #32 boosts DA chance by 5% ~ 14.5%
 * Same assumption as above: FuMO boosts DA chance by 1.5% ~ 11%

- ASTRE:
 * Using comparative method, K-Test No.6a+b & No.15 suggests a modifier of +10% DA chance for #32 radar (error +/- 3.8%). This is consistent with our finding from this exp't. For future exp't I will assume this value (+10%) as #32's DA chance modifier.
 * K-Test No.14 suggests a modifier of +6% DA chance for #33 radar (error +/- 5.2%). This is within our test results above (0~6%). For future exp't I'll keep using this value (+6% DA chance) as #33's DA chance modifier. Due to limited reports & smaller samples size, I will be more cautious on this assumption.
 * [RM%] (Radar Modifier) calculated in estimation were ~4% lower than those from comparative study. The actual cause is unknown. However, a good hypothesis is that the base DA chance for CA(V) is not 50%, but rather 45%~47%.