I was playing around trying to see what triggers rep changes ^{goto 1}. I know that there is a /reputation page, and Nick has reiterated many times that triggering a recalc is expensive (always emphasized).

I used to get rep differences between current and /reputation in the tens, which in my first 60 days at Stack Overflow happens often. I never let it go too far, since I recalc often. But since the last week or two, or since the latest global recalc that spawned the grammatically dubious floating top banner, it seems to have stabilised and at times I suspect auto-corrects itself.

Q1: Has any changes been made to make reputation track closer to the true rep (if we take result of recalc to be "true")?
Q2: Are there more points where mini-recalcs or adjustments occur?
Q3: If rep reversal is immediate in the test case to follow, why doesn't either the increase or decrease update the summary total? i.e the header had 70 v Today, the list had 200 + 0 rep votes (post rep-cap)

^1: Okay what I did was to go to waffles' (Sam's) profile, and open up 17* (or so) answers that I have not voted on. Noting that Sam has 70 rep today, I upvoted all 17* and watched the rep reach the daily limit. Then I removed the upvotes half-expecting the rep to stay (they got removed).

I was playing around trying to see what triggers rep changes ^{gosub 1}. I know that there is a /reputation page, and Nick has reiterated many times that triggering a recalc is expensive (always emphasized).

I used to get rep differences between current and /reputation in the tens, which in my first 60 days at Stack Overflow happens often. I never let it go too far, since I recalc often. But since the last week or two, or since the latest global recalc that spawned the grammatically dubious floating top banner, it seems to have stabilised and at times I suspect auto-corrects itself.

Q1: Has any changes been made to make reputation track closer to the true rep (if we take result of recalc to be "true")?
Q2: Are there more points where mini-recalcs or adjustments occur?
Q3: If rep reversal is immediate in the test case to follow, why doesn't either the increase or decrease update the summary total? i.e the header had 70 v Today, the list had 200 + 0 rep votes (post rep-cap)

^1: Okay what I did was to go to waffles' (Sam's) profile, and open up 17* (or so) answers that I have not voted on. Noting that Sam has 70 rep today, I upvoted all 17* and watched the rep reach the daily limit. Then I removed the upvotes half-expecting the rep to stay (they got removed)._return

I was playing around trying to see what triggers rep changes ^{goto 1}. I know that there is a /reputation page, and Nick has reiterated many times that triggering a recalc is expensive (always emphasized).

I used to get rep differences between current and /reputation in the tens, which in my first 60 days at Stackoverflow happens often. I never let it go too far, since I recalc often. But since the last week or two, or since the latest global recalc that spawned the grammatically dubious floating top banner, it seems to have stabilised and at times I suspect auto-corrects itself.

Q1: Has any changes been made to make reputation track closer to the true rep (if we take result of recalc to be "true")?
  Q2: Are there more points where mini-recalcs or adjustments occur?
  Q3: If rep reversal is immediate in the test case to follow, why doesn't either the increase or decrease update the summary total? i.e the header had 70 v Today, the list had 200 + 0 rep votes (post rep-cap)
 

I was playing around trying to see what triggers rep changes ^{goto 1}. I know that there is a /reputation page, and Nick has reiterated many times that triggering a recalc is expensive (always emphasized).

I used to get rep differences between current and /reputation in the tens, which in my first 60 days at Stack Overflow happens often. I never let it go too far, since I recalc often. But since the last week or two, or since the latest global recalc that spawned the grammatically dubious floating top banner, it seems to have stabilised and at times I suspect auto-corrects itself.

Q1: Has any changes been made to make reputation track closer to the true rep (if we take result of recalc to be "true")? 
Q2: Are there more points where mini-recalcs or adjustments occur? 
Q3: If rep reversal is immediate in the test case to follow, why doesn't either the increase or decrease update the summary total? i.e the header had 70 v Today, the list had 200 + 0 rep votes (post rep-cap)

^1: Okay what I did was to go to waffles' (Sam's) profile, and open up 17* (or so) answers that I have not voted on. Noting that Sam has 70 rep today, I upvoted all 17* and watched the rep reach the daily limit. Then I removed the upvotes half-expecting the rep to stay (they got removed).