A Final MatchBack analysis for General Election 2014

16,845 42nd LD Precinct Votes "left on the table" from those precincts with over 300 'not voting'. 32K voters registered in the 42nd LD did not participate.  Click on chart too enlarge.

Statistics below paint a dismal picture of participation in the 2014 General Election. 48,205 registered Whatcom County voters did not participate in the 2014 General Election including 31,938 from the LD 42nd where this year's critical races were held.  Arguably, many of the top 30 'not voting'  precincts shown above would have voted for Democrats. R Code here.
The matchback totals include 1102 duplicates registration numbers which are caused by voters attempting to "cure" their undeliverable or challenged ballots. Those first tries become void unless "cured" before certification. Undeliverable ("return to sender") is the largest category of ballot loss, follow by "Too Late" and "No signature match". But no challenge or ballot delivery issue creates as much voter loss as those that simply don't cast their ballots:

length(unique(duplicated$RegistrationNumber))
[1] 1102


count(MB_2014_Consolidated_status,"AVReturnStatus")
  AVReturnStatus  freq
1     Not Voted   48205
2     Challenged   695
3           Good 76089
4  Undeliverable  2457
5           Void   992

count(MB_2014_Consolidated_status,"AVReturnChallenge")
           AVReturnChallenge   freq
1                            125286
2                 Electronic      8
3                ID Required      9
4  Letter doesn't match env.     28
5                  No Ballot      5
6               No Signature     55
7         No Signature Match    263
8          Power of Attorney      1
9     Second Ballot Received     21
10                  Too Late    293
11             Undeliverable   2457
12   Wrong Voter's Signature     10
13              Z - Deceased      1
14                 Z - Other      1

Among those that did cast their ballots, these were the top 20 precincts:

arrange(data.frame(with(MB_2014_Consolidated_Good,(table(PrecinctID)))),desc(Freq))
    PrecinctID Freq
1          601  743
2          201  730
3          140  729
4          208  715
5          169  714
6          222  698
7          250  672
8          127  665
9          213  662
10         134  653
11         145  648
12         144  647
13         146  644
14         609  643
15         610  637
16         126  634
17         181  629
18         168  624
19         182  618
20         151  616

...

These precincts had the greatest number of votes not cast:

 arrange(subset(Test,PrecinctID %in% ld42$precincts42),desc(NoVote))

PrecinctID	No Vote	Challenged	Good	Undeliverable	Void	Undel_Void	PCT_Good	PCT_Undel_Void
133	606	7	361	14	51	65	35.0	1.4
137	568	15	366	7	50	57	36.9	0.7
229	528	4	488	23	7	30	46.7	2.2
201	521	4	730	18	7	25	57.2	1.4
206	491	6	515	22	13	35	49.5	2.1
208	469	4	715	26	8	34	58.7	2.1
108	449	5	605	3	7	10	56.9	0.3
225	448	7	482	33	12	45	49.4	3.4
148	436	2	487	11	3	14	52.0	1.2
101	421	7	476	9	4	13	52.3	1.0
302	416	7	608	14	7	21	58.2	1.3
505	414	1	575	16	7	23	56.8	1.6
211	398	6	481	7	5	12	54.0	0.8
142	395	3	373	11	11	22	47.2	1.4
220	388	4	383	20	6	26	48.1	2.5
140	385	6	729	3	4	7	65.0	0.3
508	383	4	518	7	8	15	56.6	0.8
136	373	3	459	7	15	22	53.7	0.8
166	370	4	562	7	3	10	59.7	0.7
135	369	6	458	5	7	12	54.6	0.6
506	368	0	486	8	3	11	56.2	0.9
134	362	4	653	4	9	13	63.5	0.4
126	355	2	634	9	9	18	63.0	0.9
146	354	7	644	3	9	12	63.8	0.3
203	350	1	520	15	9	24	58.2	1.7
106	348	3	422	5	0	5	54.5	0.6
204	344	1	225	13	4	17	38.4	2.2
301	343	9	600	14	10	24	62.0	1.4
219	337	5	444	8	7	15	55.8	1.0
233	337	2	489	11	7	18	57.9	1.3