Reading LP and match data

01 / RANK PROGRESS

A 30-day LP chart is not a log of every match

Start with the observation times, not the slope of the line. ANALOL saves rank and LP when rank data is successfully retrieved. It does not reconstruct LP before tracking began from match outcomes or invent values for days with no observations.

Suppose the observations are Gold II 40 LP on Monday and Gold II 68 LP on Thursday. The observed interval change is +28 LP. If three matches occurred between them, that does not mean each match was worth +9.3 LP. Outcomes alone do not establish each match's LP change or rule out non-match adjustments.

A gap can be an honest statement of uncertainty

Distinguish an interval where the same rank state was confirmed from one where a change occurred at an unknown time. A dotted line does not verify LP at every point along it. Across rank changes, read tier and division as well as LP, and do not treat resets or unsupported transitions as normal match gains.

02 / REVIEW TO ACTION

Review one decision before declaring why you lost

A high KDA does not prove you made the best decisions. Low CS can reflect missed waves, grouping with teammates, game length or role. A statistic helps you choose what to investigate; it does not establish the cause on its own.

1. Separate outcome from context

   First note outcome, role, champion and duration. Do not judge support CS and mid-lane CS against the same reference.

2. Choose one moment to inspect

   For example, inspect a death before an important objective. In the replay, check available information, allied positions, unspent gold and an exit route. Avoid using the final outcome as if all later information had been available at the time.

3. Make the next action testable

   Instead of 'play better', try 'check allied positions and an exit route before entering an unwarded choke'. Choose something you can perform and verify. Track whether you did it separately from whether you won.

Team gold leads and objective captures are not the same as one player's contribution. Use the numbers to broaden your own options, not to blame teammates or judge their character.

03 / SMALL SAMPLES

The aggregation changes what the number means

For two games of 10/1/0 and 0/9/0, the arithmetic mean of individual KDAs is 5.00, but aggregate KDA is (10+0) / (1+9) = 1.00. ANALOL uses the latter for sample-wide KDA. Do not mix an average of ratios with a ratio of totals.

Three wins in four games is a 75% win rate. One more loss brings it to 60%. Large changes in small samples are not proof that ability suddenly changed. Check role, champion, patch and time window, and always read the sample size beside the rate.

04 / WORKED EXAMPLE

Find the interval worth reviewing

In a fictional match, CS is 84 at 10:00 and 102 at 15:00. The first 15 minutes average 102/15 = 6.8 CS/min, but the 10–15-minute interval averages (102−84)/5 = 3.6 CS/min. An overall average can hide the interval worth investigating.

Established

CS increased by 18 in five minutes. The rate was lower than this player's 8.4 CS/min over the first ten minutes.

Still unknown

Whether missed last hits, a recall, grouping or enemy pressure explains it. A rate of 3.6 alone cannot establish the cause or judge performance independently of role.

Check in your replay

Inspect time spent in lane, why you left and any value gained elsewhere. If grouping achieved an objective, do not count reduced CS as the only outcome.

Next action

Check the wave and purpose of grouping before leaving. Record execution separately from winning; the goal is not to increase CS at any cost.

05 / DECISION BEFORE OUTCOME

Reconstruct the information before judging the decision

In this fictional review, a nearby ally is recalling, another ally is visible on the opposite side, and the river entrance ahead is unwarded. If the player dies afterward, the result alone does not prove that waiting was always the correct choice.

First, pause with the player's vision

Pause immediately before the decision and reason from what the player could see. Record allied positions, visible areas and the intended route. Exact unseen enemy locations, another route's safety and the time help can arrive remain uncertain without further evidence.

Compare the benefit and cost of three options

• Enter now: potentially reach the objective sooner, but accept risk while help and enemy locations are uncertain.
• Wait in a visible area: potentially gain information or help, at the cost of arriving too late.
• Pursue another objective: potentially gain value elsewhere, but concede the original objective; the new route also needs a safety check.

Time remaining, health, available abilities and wave state can change the comparison. Write down evidence and uncertainty before continuing the replay, rather than selecting one option only because you know the outcome.