Sequential Pairwise is auto-deterrent
Just mentioning something that won’t be surprising, but I mention it
because it doesn’t stand out to you unless you’re looking for
auto-deterrent methods.
Say Sequential-Pairwise (SP) is voted as rankings instead of the usual
show-of-hands. After the voting, th rankings are used to count SP.
The comparison-order agenda is determined by Topcount or Approval.
(For Approval, your ranking is counted as approving every candidate you
rank, so truncate accordingly.)
Of course candidates with higher Topcount or Approval scores are higher on
the list, earlier in the comparison order.
Say you bury a candidate (“Buried”), perceiving hir a rival to your
favorite (“BF”).
Maybe you bury hir under lots of other candidates (“Buses”).
If BUried encounters a Bus, then your order-reverse Al makes that Bus more
likely to beat & eliminate Buried. The more Buses, the more her win
probably is reduced.
But if a Bus pairbeats Buried, then you’ve only replaced Buried with Bus.
Now instead of Buried maybe being eliminated with Buried surviving & maybe
winning, now Buried is replaced with Bus, & so BF might instead be
eliminated with Bus surviving & winning.
You’ve worsened your expectation.
Your burial wasn’t a good idea.
Don’t do it.
Oops !!! I meant that higher-scoring candidates are *lower in the list, &
later in the comparison order.
On Tue, Oct 24, 2023 at 22:52 Michael Ossipoff email9648742@gmail.com
wrote:
Just mentioning something that won’t be surprising, but I mention it
because it doesn’t stand out to you unless you’re looking for
auto-deterrent methods.
Say Sequential-Pairwise (SP) is voted as rankings instead of the usual
show-of-hands. After the voting, th rankings are used to count SP.
The comparison-order agenda is determined by Topcount or Approval.
(For Approval, your ranking is counted as approving every candidate you
rank, so truncate accordingly.)
Of course candidates with higher Topcount or Approval scores are higher on
the list, earlier in the comparison order.
Say you bury a candidate (“Buried”), perceiving hir a rival to your
favorite (“BF”).
Maybe you bury hir under lots of other candidates (“Buses”).
If BUried encounters a Bus, then your order-reverse Al makes that Bus more
likely to beat & eliminate Buried. The more Buses, the more her win
probably is reduced.
But if a Bus pairbeats Buried, then you’ve only replaced Buried with Bus.
Now instead of Buried maybe being eliminated with Buried surviving & maybe
winning, now Buried is replaced with Bus, & so BF might instead be
eliminated with Bus surviving & winning.
You’ve worsened your expectation.
Your burial wasn’t a good idea.
Don’t do it.
Of course Buried—Or Bus if you’ve replaced BF with hir—might win 🏆
whether or not they encounter BF in a comparison bout.
…greatly increasing the amount by which your burial has lowered your
expectation.
On Tue, Oct 24, 2023 at 22:52 Michael Ossipoff email9648742@gmail.com
wrote:
Just mentioning something that won’t be surprising, but I mention it
because it doesn’t stand out to you unless you’re looking for
auto-deterrent methods.
Say Sequential-Pairwise (SP) is voted as rankings instead of the usual
show-of-hands. After the voting, th rankings are used to count SP.
The comparison-order agenda is determined by Topcount or Approval.
(For Approval, your ranking is counted as approving every candidate you
rank, so truncate accordingly.)
Of course candidates with higher Topcount or Approval scores are higher on
the list, earlier in the comparison order.
Say you bury a candidate (“Buried”), perceiving hir a rival to your
favorite (“BF”).
Maybe you bury hir under lots of other candidates (“Buses”).
If BUried encounters a Bus, then your order-reverse Al makes that Bus more
likely to beat & eliminate Buried. The more Buses, the more her win
probably is reduced.
But if a Bus pairbeats Buried, then you’ve only replaced Buried with Bus.
Now instead of Buried maybe being eliminated with Buried surviving & maybe
winning, now Buried is replaced with Bus, & so BF might instead be
eliminated with Bus surviving & winning.
You’ve worsened your expectation.
Your burial wasn’t a good idea.
Don’t do it.
I meant:
Of course Buried—or Bus if you’ve replaced Buried with hir—might win,
whether or not they encounter BF in a comparison bout.
You like Bus less than Buried. That’s why ranking hir over Buried was
Burial.
On Tue, Oct 24, 2023 at 23:06 Michael Ossipoff email9648742@gmail.com
wrote:
Of course Buried—Or Bus if you’ve replaced BF with hir—might win 🏆
whether or not they encounter BF in a comparison bout.
…greatly increasing the amount by which your burial has lowered your
expectation.
On Tue, Oct 24, 2023 at 22:52 Michael Ossipoff email9648742@gmail.com
wrote:
Just mentioning something that won’t be surprising, but I mention it
because it doesn’t stand out to you unless you’re looking for
auto-deterrent methods.
Say Sequential-Pairwise (SP) is voted as rankings instead of the usual
show-of-hands. After the voting, th rankings are used to count SP.
The comparison-order agenda is determined by Topcount or Approval.
(For Approval, your ranking is counted as approving every candidate you
rank, so truncate accordingly.)
Of course candidates with higher Topcount or Approval scores are higher
on the list, earlier in the comparison order.
Say you bury a candidate (“Buried”), perceiving hir a rival to your
favorite (“BF”).
Maybe you bury hir under lots of other candidates (“Buses”).
If BUried encounters a Bus, then your order-reverse Al makes that Bus
more likely to beat & eliminate Buried. The more Buses, the more her win
probably is reduced.
But if a Bus pairbeats Buried, then you’ve only replaced Buried with Bus.
Now instead of Buried maybe being eliminated with Buried surviving &
maybe winning, now Buried is replaced with Bus, & so BF might instead be
eliminated with Bus surviving & winning.
You’ve worsened your expectation.
Your burial wasn’t a good idea.
Don’t do it.