dailysudoku.com

ravel · Joined: 21 Apr 2006 Posts: 536

Again a multidigit deadly pattern can help a lot (one step solution).
A remark for those, who are not familiar with them: As Myth points out, whenever you have such a deadly pattern and you place one of the digits outside of it, the remaining pattern also is deadly.

wapati · Posted: Fri Aug 22, 2008 5:08 pm Post subject:

Interesting puzzle. I see two hidden URs that overlap in r6c4. They destroy each other so I look at both before eliminating.

tlanglet · Posted: Fri Aug 22, 2008 6:38 pm Post subject:

I could not solve the MUG as a one step solution even after re-reading Myth's comments. However, after a x-wing, a w-wing and another x-wing, the MUG had only one cell with an extra candidate, r5c4, Setting that cell to the extra candidate broke the puzzle open.

Ted Smile

ravel · Joined: 21 Apr 2006 Posts: 536

ravel · Joined: 21 Apr 2006 Posts: 536

Just not to be misunderstood. This MUG is not the best sample by Myth.

tlanglet · Posted: Fri Aug 22, 2008 7:19 pm Post subject:

ravel · Joined: 21 Apr 2006 Posts: 536

Posts crossed Smile

The 458's dont form a MUG.

tlanglet · Posted: Fri Aug 22, 2008 8:18 pm Post subject:

Asellus · Posted: Fri Aug 22, 2008 8:24 pm Post subject:

If I understand ravel's MUG correctly, it eliminates a bunch of <7>s and is a nice find.

This puzzle is a good one for practicing basic Medusa coloring (no "extentions" or multi-coloring required). Start with the bivalues in column 3. After a few "traps" or so the puzzle falls apart. Or, if you first apply the <4> X-Wing, basic Medusa leads quickly to a "wrap."

Earl · Joined: 30 May 2007 Posts: 677 Location: Victoria, KS

After the x-wing (4), an xy-chain eliminates the 4 in R2C3 and solves the puzzle. Testimony to the quick brute force of xy-chains.

Earl

Marty R. · Posted: Sat Aug 23, 2008 1:29 am Post subject:

I used a hidden UR on 58, three ERs, X-Wing on 4, Finned XY-Wing on 794 (just a forcing chain with a fancy name), then the actual 794 XY-Wing with pincer coloring finished it. Medusa looked like a good possibility but I just decided on the Finned deal instead because I'm too Medusa-dependent.

wapati · Posted: Sat Aug 23, 2008 4:43 am Post subject:

Kdelle · Joined: 20 Mar 2008 Posts: 59 Location: Hudson, NH

ravel · Joined: 21 Apr 2006 Posts: 536

The shortest i could find (7 cells):

tlanglet · Posted: Sun Aug 24, 2008 1:26 am Post subject:

daj95376 · Joined: 23 Aug 2008 Posts: 3854

ravel · Joined: 21 Apr 2006 Posts: 536

There are several ways use deadly patterns.

The most common way is to look at the extra candidates (like in URs type 1, 2 and 3). One of them must be true, otherwise you are left with the DP, which means, that the puzzle has no or multiple solutions.
If this does not lead to an elimination directly, you also can try to "transport" them to find one, just like with double or multiple forcing chains. This is, what Danny showed above. Both extra candidates lead to a 7 in column 6 of box 2. Or you can use this relationship in longer chains or AIC's, e.g. 2 extra candidates mean a strong link between the digits - here (9)r1c1=(6)r3c1. This gives an AIC (7=9)r1c6-(9)r1c1=(6)r3c1-(6=7)r3c6 with the same result, that one of r13c6 must be 7.

Then you can look at strong links of DP digits (like in URs type 4 and 6 or "hidden" URs). If placing a digit would force a deadly solution pattern (like 1-2-1-2 in UR cells) because of these strong links, you can eliminate it. This is generally true (not restricted to strong links and DP digits).

Another way is to use the fact, that a deadly pattern can only be destroyed, if at least one of the DP digits is outside in the DP units. E.g. you can look at a UR type 4 like this: